SCANNED. CLEANED.

AND COMPILED BY
mEKENDAMA OF �OLOMBIA
AND �LOBAL �OUNTY 8UNTERS

The
[ftAI�IBIIIl
Guide to the
Modern Benoni

John Watson
Programs used: VueScan, Photoshop, ScanKromsator,
SnagIt, Acrobat

E-mail: kerezhma@gmail.com

Intended solely for the use of trading chess materials

with select members of the GBH chess club

mAI��IBIITI
First published in the UK by Gambit Publications Ltd 200 1

Copyright © John Watson 2001

The right of John Watson to be identified as the author of this work has been as­
serted in accordance with the Copyrights, Designs and Patents Act 1 988.

All rights reserved. This book is sold subject to the condition that it shall not, by
way of trade or otherwise, be lent, re-sold, hired out or otherwise circulated in
any form of binding or cover other than that in which it is published and without a
similar condition including this condition being imposed on the subsequent pur­
chaser.
A copy of the British Library Cataloguing in Publication data is available from
the British Library.

ISBN 1 90 1 983 23 4

DISTRIBUTION:
Worldwide (except USA): Central Books Ltd, 99 Wallis Rd, London E9 5LN. Tel
+44 (0)20 8986 4854 Fax +44 (0)20 8533 582 1 .
E-mail: orders@Centralbooks.com
USA: BHB International, Inc., 4 1 Monroe Turnpike, Trumbull, CT 066 1 1 , USA.

For all other enquiries (including a full list of all Gambit Chess titles) please
contact the publishers, Gambit Publications Ltd, 69 Masbro Rd, Kensington,
London W 1 4 OLS.
Fax +44 (0)20 737 1 1 477. E-mail Murray@gambitchess.freeserve.co.uk
Or visit the GAMBIT web site at http://www.gambitbooks.com

Edited by Graham Burgess

Typeset by Petra Nunn
Printed in Great Britain by The Bath Press, Bath, Somerset

10 9 8 7 6 5 4 3 2 1

Gambit Publications Ltd

Managing Director: GM Murray Chandler
Chess Director: GM John Nunn
Editorial Director: FM Graham Burgess
German Editor: WFM Petra Nunn
Contents

Symbols 4
Acknowledgements 4
Foreword 5
Bibliography 8

I Introduction and Early Moves 9

2 Systems with i..f4 22
3 Systems with ..tg5 44
4 T he Knight's Tour Variation 59
5 Pawn-Storm Systems 72
6 Fianchetto Systems with g3 110
7 Systems with i..d3 and tbge2 130
8 Kapengut's 7 f3 System 147
9 Modem Main Line 157
10 Classical Main Line 186

Index of Variations 206

Symbols
+ check tt team tournament
++ double check Wch world championship
# checkmate Ech European championship
!! brilliant move Wcht World Team Championship
! good move ECC European Clubs Cup
!? interesting move Ct candidates event
?! dubious move IZ interzonal event
? bad move Z zonal event
?? blunder OL olympiad
+- White is winning jr junior event
± White is much better worn women's event
;j; White is slightly better mem memorial event
= equal position rpd rapidplay game
00 unclear position corr. correspondence game
=+= Black is slightly better 1 -0 the game ends in a win for White
+ Black is much better liz-liz the game ends in a draw
-+ Black is winning 0- 1 the game ends in a win for Black
Ch championship (n) nth match game
Cht team championship (D) see next diagram

Acknowledgements
I would like to thank those who have helped me with this book, including Nick de
Firmian, Mark Quinn, Ken Case and John Donaldson. I am particularly grateful
to Graham Burgess for his numerous analytical corrections and suggestions.
Foreword

The project before you, like so many investigations in the chess world, has
proven full of surprises. As often happens with authors, I have been struck by
how theory, incredibly deep in some lines, is superficial or just wrong in many ar­
eas. We tend to assume that everything is more or less worked out in established
openings, but thllt is perhaps a confusion between the great mass of games,
which often don't signify much, and the best moves. It turns out that much of the
Modem Benoni is still unexplored territory.
This is a repertoire book for Black, which means that instead of trying to cover
the opening as a whole, Black is provided a set of suggested lines. The drawback
to such a book is its lack of comprehensiveness; its advantages are in attention to
detail and in the discipline it imposes upon the author. That is, the author is re­
sponsible for specific lines of play, and to the extent that it is practical, should not
ignore any reasonable move by White, whether or not it has been played or dis­
cussed.
My method has been to offer the reader a first, second and occasionally third
system against the most critical and popular variations. Thus if one line fails due
to a theoretical development, or if it doesn't appeal to you, there's another option
or two to look into. The majority of older or lesser variations are dealt with in the
same fashion, although there are some exceptions. In most cases, the first sug­
gested repertoire system is a more established strategy, whereas the second rep­
ertoire system tends (with very important exceptions) to be somewhat lesser
known, less theoretical, and in many cases, more fun. I have also analysed or
simply mentioned further options for Black in the notes to these proposed sys­
tems.
Given the limitations of space, my preference has been to forego chapter intro­
ductions that include diagrams of characteristic positions and schemes. I feel that
prose explanations in opening books should be linked to the examples and analy­
sis, and have therefore incorporated most of my general rules, tips, and advice
within the analysis and not in isolation. Regarding the structure of the presenta­
tion, one will note that the majority of lines end rather pleasantly for Black, re­
flecting my own advocacy. But that is an aesthetic choice of layout and by no
means indicative of a final assessment. From the analysis, it should be very clear
at which points both sides could have pursued their most promising strategies.
Thus the notes are vitally important, and I will always try to draw the reader's
6 THE GAMBIT GUIDE TO THE MODERN BENONI

attention to the critical junctures. In that context, if I think that a particular

unplayed or underestimated line is White's best chance, I will devote extra anal­
ysis to that option, which accounts for some notes being more complex than the
main line itself. In no case will I recommend a system that, played correctly,
gives White a large, unambiguous advantage, however appealing that system
may be for Black in practical play. Sometimes I give a brief analysis to indicate
why I have not used the theoretically preferred line for Black. Finally, for better
readability, I have mostly reserved the use of diagrams for intelligibility (e.g.,
before complicated branches), rather than for novelties or stunning moves.
Inevitably, I used computer engines (mostly Fritz 6, Nimzo 7.32, and Hiares
7.32) to catch blunders (numerous in my case !), monitor positions, and suggest
moves, roles which they fulfilled admirably. From previous experience, I knew
that when a truly subtle move or one requiring a long-term view was required, the
engines were seldom up to the task, even when given overnight to look, and/or
when they were allowed to look for a second- or third-best move. So I spent a
great deal of time with a physical board, pen, and notebook in hand, especially
when writing a chapter for the first time. My most original ideas of significance,
such as they are, arose from that environment. One must be careful to ignore
many of those '±'s and '+'s that the engines generate, not a few of which will be
overturned in short order if you just trust your judgement and play some moves
for the purportedly inferior side. On the other hand, by allowing these inhuman
assistants to help, I have certainly been able to include more original analysis in
this book than in any other I have written. As always, if a note has no attribution,
it comes from me, except for rare cases of a few painfully obvious moves. I have
also not been at all shy about citing game fragments by weak or unknown players
if they involve a continuation the reader may reasonably want to know about. It is
perfectly simple to interrupt the example with a suggested improvement, just as
with higher-level games. Limiting oneself to the latter is not only snobbish but
also a guarantee of decreased quality.
My attitude is biased towards Black, however objective I strive to be; and as
time goes on, readers and theoreticians will doubtless find that some of my as­
sessments are too optimistic for Black (as well as for White, but probably in
fewer cases). I have not 'hidden' any adverse assessments, and have reluctantly
weeded out attractive options that didn't pan out. Even when I have some reser­
vations about a suggested line on the basis of general impression, I have tried to
go by the verdict of my analysis rather than my intuition. Nevertheless, I would
be surprised if every one of my more ambitious and experimental second lines
survives scrutiny in the long run. I do believe that most of them will.
What is the theoretical status of the Benoni? Without question, the two sys­
tems that most threaten the viability of this opening are to be found in Chapter 9
(the 'Modem Main Line') and in Line A3 of Chapter 5 (the 'Taimanov Attack').
 FOREWORD 7

For the last 1 5 years, players of Black have struggled to find answers to the diffi­
culties posed by those set-ups. I have attempted to present specific and detailed
solutions to both and I leave it to the reader to judge whether I have succeeded.
Serious challenges are also posed by some of the i..f4 systems in Chapter 2, the
i..d3 and tLJge2 strategies of Chapter 6, and in a more positional sense, by the
Classical variation of Chapter 1 0; but in the end, Black seems well able to cope
with those approaches. The Benoni resembles other ambitious defences, in that
one strives for rich counterplay and tries, at least to some extent, to render the
question of theoretical equality moot. At the end of many variations, an author
simply has to live with a dynamic imbalance that defies proper assessment.
The Benoni was my first defence, inspired by the play of the great Mikhail Tal.
Other world champions such as Spas sky and Fischer used the Benoni only spo­
radically, but the early Kasparov had it as his main weapon. The best-known con­
temporary Benoni players and theoreticians are probably Psakhis, Kapengut,
Nunn, Suba and Topalov. A selective list of other players who have strongly con­
tributed to its theory would include Suetin, Velimirovic, de Firmian, D.Gurevich,
Sax, Ljubojevic, Matulovic, Pigusov, Lobron, Magerramov, Marin, Shabalov,
Kindermann, Y.Griinfeld, A.Schneider, Yudasin, Spraggett, Wahls, Wedberg and
Emms. My apologies to the other deserving names I have left off this list.
Enjoy this book, and have fun with your Benoni adventures !

John Watson
Bibl iogra p hy

Leading Sources
Kapengut, Albert and Gelfand, Boris; A65 Benoni; Sahovski Informator 1 996
Kapengut, Albert and Gelfand, Boris; A70 Benoni; Sahovski Informator 1 998
Kapengut, Albert; Indiskaya Zashchita, Polymia 1 984
Matanovic, Aleksandar et a1.; ECO A, 2nd Edition ('ECO'); Sahovski Informator
1 996
Psakhis, Lev; The Complete Benoni; Batsford 1 995
Schneider, Attila; Die Komplette Moderne Benoni-Verteidigung: Vols 1 -2,
Reinhold Dreier 1 997; Vol 3, Reinhold Dreier 1 998
Databases, primarily Mega Database 2000 and Corr Database 2000
(correspondence games); ChessBase 2000
Informators 1 -78
The Week in Chess 1 - 3 1 2
NIC Yearbooks 1 -55 (including all Kapengut surveys)
CBM Magazines 1 -77
ChessPublishing (Internet)

Sources Less Frequently Used

de Firmian, Nick; Fedorowicz, John; Donaldson, John; Leverett, Bruce, et al. ;
Batsford's Modern Chess Openings ('MCO'); Batsford 2000
Hebert, Jean; Modern Benoni (CD-ROM); ChessBase 2000
Norwood, David; The Modern Benoni; Cadogan 1 994
Nunn, John; The Benoni for the Tournament Player; Batsford 1 982
Vaisser, Anatoli; Beating the King's Indian and Benoni; Batsford 1 997
Nunn, John; Burgess, Graham; Emms, John; Gallagher, Joe; Nunn's Chess
Openings ('NCO'); GambitlEveryman 1 999
Ward, Chris; Modern Benoni (video); Grandmaster Video 1 999
Watson, John; Taimanov Attack and Knight's Tour Benoni; Chess Enterprises
1 985
1 I ntrod uction a nd
Ea rly M oves
The Benoni is first and foremost an ac­
tive defence. Opening books often
claim that the opening they advocate is
'for the attacking player' , 'adventur­
ous' , 'not for the faint-hearted' , and
the like. But the Modem Benoni, cor­
rectly played, truly fits such character­
izations. There is no ' Symmetrical
Benoni' , no system that could be fully
described as 'The Positional Line ' ,
and only the rarest instance o f an
early exchange of queens. Probably no
other respectable opening requires as
many pawn and exchange sacrifices to
achieve a good game, and proper Ben­
oni play includes the most extreme ex­
amples of competing attacks and
counterattacks.
John Nunn, describing the Benoni,
says "Black relies fairly heavily on
tactical resources to vindicate his
opening play. Usually there will come
a moment when Black will have to
continue tactically to justify his play,
for otherwise his pieces will be pushed
back from their active squares and he
will be reduced to permanent passiv­
ity." The remarkable thing is that such
tactical resources persistently appear
for Black, as first shown by Mikhail
Tal, the hero of Benoni players every­
where.
Despite the considerable number of
international masters and grandmasters
who use it, the Modem Benoni is still
considered marginal by many contem­
porary players. To the extent that their
scepticism is founded upon general
considerations, I think they might ar­
gue that White controls more space,
and that if White can suppress Black's
principal freeing moves ( ... b5, and
sometimes ...f5), he will have plenty
of time to organize an attack of his
own. The philosophy behind the latter
point is that possession of greater
space allows one to transfer pieces to
the attack more quickly than the oppo­
nent can. In certain Benoni positions,
White also has the two bishops to as­
sist him (usually after an exchange of
bishop for knight involving ...it.g4 and
... it.xf3 or . . .b6, ... .ta6 and . . .it.xc4).
Modem practice provides a number
of counterexamples to this way of
thinking. Numerous black defences
compensate for a lack of space by the
dynamic elasticity of their pawn­
structures. A leading example is the
Open Sicilian Defence, in which White
controls more space (generally four
ranks to three), especially in the ... e6
variations such as the Scheveningen
and many lines of the Najdorf, but also
10 THE GAMBIT GUIDE TO THE MODERN BENONI

in vanatIOns such as the Maroczy

Bind, PaulseniKan, etc. Black' s free­
ing moves such as ...b5 and ... d5 can w
sometimes be suppressed, but structural
considerations and the latent activity
of Black's pieces ensure a balance.
Similar concepts are expressed in
defences as varied as the King ' s In­
dian, Modem, French, Nimzo-Indian
and others. In many Benoni positions,
even if White manages to avoid the
kind of tactical resource mentioned
above, his control of more space by no
means guarantees that he can improve undermining the heart of Black' s posi­
his position. If White undertakes fur­ tion, his pawn on d6. A successful e5
ther space-gaining moves such as g4, can both restrict the opponent' s forces
as, or b4, they often prove to be weak­ and open up lines for White's pieces.
ening and backfire. In the meantime, Black' s normal counter-strategy is to
Black's possibility of breaking down restrain the advance of the e-pawn by,
the centre and his threatened expansion for example, ... tbbd7, ... :te8, ... "fIc7,
on the queenside are difficult to sup­ etc. Only when that is achieved can he
press indefinitely. Moreover, many safely undertake operations else­
Benoni endgames are equal or even fa­ where. Here are two typical examples:
vourable to Black, given his compact
pawn-structure. So there is a certain
onus on White to act in the middle­
game if he is to achieve anything. w
It would be impossible to illustrate
even a fraction of typical positions
and tactics that arise from the Benoni.
My object has been to discuss these
when they arise in concrete positions
throughout the book. Nevertheless,
let me comment upon just a few
prototypical structures and strategic
themes for those who are new to this
opening.
The most fundamental strategy for A position from Chapter 5. White
White when he has the kind of struc­ employs his f-pawn and knight to en­
ture featured in the following dia­ force e5; Black uses his d-pawn, rook
gram is to enforce the advance e4-e5, and knight to prevent it.
 INTRODUCTION AND EARLY MOVES 11

the difficulties that might arise, would

like to restrain . . .bS by a4, counting
w upon his knight on c3 and sometimes a
bishop on e2 or d3 to help out in this
task. In fact, one example of this is the
last diagram, in which . . .bS is held
back by the pawn on a4 and knight on
c3. Black may add support to the ...bS
thrust by moves such as . . ..l:.ab8 and
. . . lUe8-c7, whereas White may play
moves like .ie2 and/or �3 in an at­
tempt to prevent it.

From Chapter 9. Here White keeps

his f-pawn back, but tries to support
the eS advance with his bishop on f4 B
and his rook on e l . Two other aspects
of this position are worth noting:
a) The presence of the queen on e7,
which makes Black vulnerable to the
tempo-winning d6 if he should re­
spond to eS.by . . .dxeS .
b) Black has played ... �g4 and ex­
changed the bishop for a white knight
on f3. This removes a piece which
would normally support the eS ad­
vance, while conceding the two bish­ This is an example from Chapter
ops. One could argue, however, that 10. White is doing his best to hold
knights are the equal of bishops in po­ down ...bS with a pawn, knight, bishop
sitions characterized by mutual re­ and queen. If Black is equally stubborn,
straint. he might continue by ...'iVc8, . . .l:tab8,
One of Black's standard plans in and . . .'iVb7, just to enforce . . .bS. Of
most variations is to activate his queen­ course, both sides have other things to
side majority by ...bS. This can be fol­ do as well. By the way, this position
lowed by moves like ... c4, ... b4 and arose after Black played . . .b6 and
...c3, which will drive back White's . . ..ia6, after which the bishop on a6
pieces and sometimes extend the in­ took a knight on c4. See below for a
fluence of the g7-bishop. The combi­ discussion of both the lUf3-d2-c4 ma­
nation of . . . c4 and . . .lUcs can exert noeuvre and Black's desire to ex­
pressure on important central squares change his light-squared bishop for
such as e4 and d3. White, anticipating that knight.
12 THE GAMBIT GUIDE TO THE MODERN BENONI
w Black need not always prevent it di­
The e5 versus . . . b5 struggle fre­
quently leads to a violent clash of
forces. In this position from Chapter 7,
White is about to tear Black apart with
e5, but ...b4 is also corning, to drive
White's pieces back away from the
centre and initiate a counterattack.
Another idea here is that e5 can be met
by ...dxe5 and then an exchange sacri­
fice on e5 (i.e., ....l::.x e5), after which
Black's pieces become extremely ac­
tive and he controls the dark squares.
There are other standard ideas asso­
ciated with these e5 and ... b5 themes.
One of these arises when White sim­
ply allows ...b5, but then stops the ad­
vance of the queenside majority by
.l::.b l and b4. A major idea in that case is
that the move ... c4 concedes to White
the use of d4 for his pieces, whereas
Black's knight is denied access to c5.
On the other hand, a move like ...cxb4
in response to b4 will often help Black
to overrun the queenside. In another
scenario, White can wait for ...b5 and
then play a4, in order to respond to ...b4
by ttJbl or ttJdl . His idea is to occupy
c4, but the waste of time involved can
make White vulnerable to central tac­
tics. Regarding White's e5 break,
rectly by massing his forces against
that square, but will frequently take
prophylactic measures to ensure that
e5 exposes White's d-pawn. Thus, a
queen or rook on dS with knights on f6
and/or c7 can ensure that White' s e5
advance loses the d-pawn. Even if d5
is sufficiently protected after e5, the
response ... dxe5 will often allow an at­
tack on the now vulnerable d-pawn
by, say, ... .i.b7 and ....l::.adS.
A manoeuvre characteristic of ev­
ery system except those of Chapters 7
and S is ttJf3-d2-c4 for White. One
might think that this is too time-con­
suming, especially since the knight on
f3 is already watching over d4 and e5.
But there are several reasons for this
sojourn. For one thing, the d6-square
is attacked, and .i.f4 will attack it a
second time, so Black must keep two
of his own pieces ready for defence of
that square. This means that standard
moves such as ... ttJbd7, . . .ttJfd7, and
. .. .i.d7 are sometimes unplayable, be­
cause they cut off the queen's protec­
tion of d6. The c4-knight can also be
very strong in conjunction with f4 and
e5, not least because it supports a
timely d6 after ... dxe5.
A knight on c4 tends to be so effec­
tive that if Black cannot drive it away
by ... b5, he will usually try to ex­
change it. In this example from Chap­
ter 6 (see diagram on/allowing page),
he has just challenged the c4-knight
by 1 3 . . .ttJb6, yet after 14 ttJa3, Black
seems to have wasted time, since a5
 INTRODUCTION AND EARLY MOVES
w larly difficult to post effectively, but
will expel Black's knight and then
tbc4 follows anyway. Nevertheless,
Black can respond by 14 ... i.d7 15 a5
tbc8 1 6 tbc4 i.b5, once again intend­
ing to trade off the knight, which has
no particularly good place to go. As
mentioned earlier, a similar bishop­
for-knight trade on c4 can arise after
... b6 and ... .i.a6, or Black can act pro­
phylactically, by an early i.g4 and
• ...
... i.xf3, to stop the f3-knight from
even starting out on its journey.
This idea of exchanging off Black's
c8-bishop recurs throughout the Ben­
oni. Superficially, one might think that
this is his 'good' bishop (unrestricted
by his centre pawns). But in fact, the
c8-bishop frequently has nowhere use­
ful to go, since white pawns on d5 and
e4 prevent ...i.e6 or ... i.f5, whereas
g4 is often covered by a bishop on e2
or a pawn on f3 or h3. On the queen­
side, this same bishop can look silly
on b7 'attacking' the super-solid pawn
on d5 (an exception occurs when White
succeeds in playing e5 and his d-pawn
becomes weak). At any rate, Black
should be aware that his light-squared
13
bishop is a problem piece. White's
own light-squared bishop can be simi­
for the traditional reason that his cen­
tre pawns tend to be on light squares.
There are many other themes and
trade-offs that one will encounter re­
peatedly throughout this book; for ex­
ample:
a) the effects of Black's freeing
move ... f5;
b) the consequences of expansion
by ...c4 without ... b5 ;
c) the conditions for pawn sacri­
fices involving e5 or ...b5 ;
d) the weakening of Black's king­
side by ...h6 and ...g5 to win the two
bishops;
e) White's attack by f4-f5;
f) White's attempted expansion by
g4; and
g) the special problems introduced
by the move .i.g5.
These will be examined in context.
One last issue, however, deserves dis­
cussion at this point, since it arises in
nearly every Benoni variation, and may
well confuse those new to the opening.
Returning to the queenside struggle,
White will often play a5 in response to
... a6, so that he can capture en passant
should Black play ...b5 . Black can pre­
vent this by playing ... b6, then ... a6,
and finally ...b5, but that is rather slow.
So he usually prefers to play ...b5 di­
rectly and allow White's capture on
b6. Then we have a situation in which
Black has the b-file with a target on
b2, whereas White has the a-file with a
target on a6. Here are two of literally
hundreds of such positions:
14 THE GAMBIT GUIDE TO THE MODERN BENONI
From Chapter S . Black's pressure
down the b-file counteracts White's
space and bishops. After White de­
fends the b-pawn, Black has moves
like .. J:tb3 and . . .'�jbS-d4 to try to keep
the initiative.
w izations will help you orientate your­
A knight on b6 doesn't exert direct
pressure as the rook did in the last dia­
gram, but it can still be effective. In
this position from Chapter 6, White
might be worried about ...fS and opt
for h3, but after .. .'!i:JeS, Black is ready
for . . .'�jbc4. White's attempt to prevent
this by b3 would be well met by . . . c4 !,
threatening infiltration on d3 .
These last two positions are fairly
level, but I'll go out on a limb and state
that in a majority of cases, Black gets
more out of the b-file than White does
from the a-file. This is partly because
the queenside is his natural sphere of
activity, and partly because the b4-
and bS-squares are so helpful in acti­
vating his pieces. Thus Black nor­
mally shouldn't fear as, and in fact, he
sometimes provokes it. Of course,
there are exceptions; e.g., when White
restricts Black's game with a knight or
bishop on as, or when the a-pawn is
particularly weak. Also, a position of
mutual zugzwang on the queenside
generally favours White, who gains a
freer hand on the rest of the board.
Fortunately for Black, that kind of sta­
sis is relatively infrequent.
I hope that these general character­
self as you delve into the particulars of
this book. That said, let's move on to
the obligatory task of examining some
early deviations and questions of
move-order.
1 d4 tLlf6 2 c4 c5 3 d5 e6
The first move of the Modem Ben­
oni. One might argue that the Benoni
proper doesn't begin until after 4...exdS
S cxdS d6 (or even after 6. . . g6). But on
the way there White has a few early al­
ternatives, some designed to transpose
at his discretion. The theoretically
most important of these is certainly 4
tLlf3 exdS S cxdS d6 6 e4 (Line C 1 ), in
which White either foregoes or delays
tLlc3. Fortunately, although they can
 INTRODUCTION AND EARLY MOVES
hardly be disadvantageous, the other
deviations tend to be passive and
shouldn't overly worry Black:
A: 4 dxe6 IS
B: 4 g3 16
C: 4 ttJf3 17
D: 4 ttJc3 19
Throughout the book, I will empha­
size the fact that there is a significant
distinction between Benoni systems in
which White is already committed to
the move ttJf3 and those in which he
delays or foregoes that move. This dif­
ference is particularly important in
practice, because many d-pawn de­
fenders will use the move-order 1 d4
ttJf6 2 c4 e6. Then, after 3 ttJc3, they
will choose the Nimzo-Indian (3 ...i.b4)
or the Queen's Gambit Declined (3 ...dS)
rather than enter a Modem Benoni by
3 ...cS 4 dS exdS S cxdS d6. But these
same players are often perfectly will­
ing to play the Benoni after 3 ttJf3 cS 4
dS exdS S cxdS d6, precisely b�cause
White has already played ttJf3. In that
case, Black does not have to deal with
the white systems of Chapters S, 7 and
8, and also gains some flexibility
against other systems, a prime exam­
ple being his ability to delay ... i.g7 in
favour of . . . a6 in certain variations of
the Modem Main Line of Chapter 9. I
will often refer to the differences be­
tween systems with and without ttJf3
as we move along.
A}
4 dxe6 fxe6 (D)
To me, 4 . . .dxe6 S 'iixd8+ �xd8
looks playable, but it's not surprising
15
that Black wants something more in­
teresting.
w
Although 4 dxe6 is barely men­
tioned by theory, it can hardly be that
bad. Black gets an extra centre pawn,
and White gains the opportunity to ex­
ert pressure down the d-file. After
4 .. .fxe6, we see:
a) S i.gS dS ! ? (S ...ttJc6 is a solid
approach; e.g., 6 ttJc3 i.e7 7 e4 0-0 8
ttJf3 d6 =) 6 e4 ! ? (more ambitious than
6 e3 i.e7 { 6 . . .ttJc6 is also fine } 7 ttJc3
d4 ! ? { again, there is nothing wrong
with 7 . . .ttJc6 } 8 exd4 cxd4 9 i.xf6
i.xf6 10 ttJe4, P'Larsen-OJakobsen,
Herlev 1 997, when 10 ...0-0 1 1 i.d3 eS
1 2 ttJf3 i.g4 = could follow) 6 ...h6! (a
bold pawn sacrifice; 6. . .i.e7 7 cxdS
exdS 8 exdS 'iVxdS is equal but dull) 7
i.xf6 (after 7 i.h4 gS ! 8 i.g3 ttJxe4 9
'iVhS+ �e7 White lacks compensa­
tion) 7 .. :ii'xf6 8 exdS exdS 9 cxdS ! ? (9
li'e2+ �d8 ! 10 ttJc3 d4 1 1 ttJdS 'iVg6
12 0-0-0 i.d6 1 3 ttJf3 l:.e8 14 'iid2
ttJc6 gives Black a safe king-position
and good piece-play) 9 ...i.d6 10 i.bS+
ttJd7 1 1 ttJc3 0-0 (Black has plenty of
16 THE GAMBIT GUIDE TO THE MODERN BENONI
activity for a pawn) 12 tZ:\f3 tZ:\eS 1 3
.te2 tZ:\xf3+ 1 4 .txf3 .td7 I S 'iVc2 ( 1 S
tZ:\e4 1be8 1 6 O-O? lIxe4 ! 1 7 .txe4
�f4) I S . . .,Uae8+ 16 .te4 .tfS 17 f3?
( 1 7 0-0-0 .txe4 1 8 tZ:\xe4 'iVf4+ 1 9
tZ:\d2 'it'xf2 =+=) 1 7 . . :ii'h4+ 1 8 �e2 bS !
1 9 g3 'ii'hs 20 �d2 .txe4 2 1 tZ:\xe4
'iVxdS+ 22 �e2 'i'hs 23 g4 'ikh3 24
:afl :xf3 0- 1 Foguelman-Mecking,
Buenos Aires 1 967. This is the most
famous game with 4 dxe6.
b) S tZ:\c3 dS (S ... tZ:\c6 6 e4 .te7 7
tZ:\f3 0-0 =; S ....te7 6 g3 dS 7 .tg2 d4 8
tZ:\e4 0-0 9 tZ:\xf6+ .txf6 10 tZ:\f3 .td7
1 1 h4? .tc6 1 2 'it'c2 eS =+= 1 3 tZ:\gS?? d3
-+ Morgner-Nikitin, Dresden 1 993) 6
cxdS exdS 7 .tgS (7 e3 tZ:\c6 8 tZ:\f3
.te7 =) 7 ...d4 8 tZ:\e4 �b6! (8 ... .te7 = )
9 tZ:\xf6+ gxf6 10 .te l .tfS =+=; Black
can follow up with the moves ...tZ:\c6
and ...0-0-0.
c) S g3 is probably the most fre­
quent move, although it seems to be no
better than the others: S ...tZ:\c6 6 .tg2
dS (again, Black can delay a central
commitment; e.g., 6 . . ..te7 7 tZ:\f3 0-0
8 0-0 dS 9 e3? ! l:[e8 1 0 tZ:\c3 d4 1 1
exd4 cxd4 1 2 tZ:\a4 eS =+= Padrak-Sim­
antsev, Polanica Zdroj 1 999) 7 tZ:\h3 ! ?
d4 8 tZ:\f4 .td6 9 tZ:\d3 h 6 1 0 e4 eS (re­
sembling a King's Indian Defence
with colours reversed) 1 1 f4 ( 1 1 0-0
i.e6 1 2 tZ:\d2 0-0 1 3 f4 exf4 1 4 gxf4
tZ:\g4!) 1 l . ..i.g4 1 2 .tf3 hS 1 3 tZ:\d2
�d7 1 4 fS ( 1 4 0-0 h4 !) 14 ... g6 ! I S h3
( 1 S fxg6 0-0-0 with the ideas ....:tdg8
and ... h4) I S ...i.xf3 1 6 �xf3 gxfS 1 7
exfS ( 1 7 �xfS ? 'it'xfS 1 8 exfS e4 1 9
tZ:\f4 i.xf4 20 gxf4 tZ:\b4) 1 7 ...0-0-0 1 8
tZ:\e4 tZ:\xe4 1 9 �xe4 lldg8 20 �f2 tZ:\e7
2 1 tZ:\xeS �xfS+ 22 �xfS+ tZ:\xfS 23
.tf4 (23 g4 tZ:\e3) 23 . . .h4! + Mocha­
lov-Tseshkovsky, Minsk 1 982.
B)
4 g3 exd5 5 cxd5
This will almost always transpose
to Chapter 6 after S ...d6 6 .tg2 g6 7
tZ:\f3 .tg7 8 0-0 0-0 9 tZ:\c3, which I
feel is fine for Black. Nevertheless,
some very good players have opted
for.. .
5...b5!? (D)
w
As this is not strictly speaking a
repertoire move, I will give it only a
superficial look.
6 .tg2
6 a4 b4 7 f3 ! ? g6 8 e4 .tg7 9 tZ:\d2
0-0 1 0 tZ:\c4 d6 1 1 .tf4 .ta6! = Korch­
noi-Timman, Tilburg 1 987. Then 1 2
tZ:\xd6? would fail to 1 2 ...tZ:\hS.
6 d6
••.
6....tb7 7 e4 d6 is also played.
7 tZ:\f3
An intriguing gambit is 7 b4 ! ? cxb4
(7 ....tb7 ! Timman) 8 a3 bxa3 9 tZ:\xa3.
A good answer is 9 ....td7 10 tZ:\f3 as
1 1 0-0 .te7 1 2 tZ:\d4 b4 1 3 tZ:\c4 :a6,
 INTRODUCTION AND EARLY MOVES
as i n Hulak-Podlesnik, Portoroz 1 996.
The idea 7 a4 b4 8 ttJd2 was answered
by 8 . . .ttJbd7 ! ? (to be ready to play
...ttJb6 in reply to ttJc4) 9 b3 g6 1 0
i.b2 i.g7 1 1 ttJc4 ttJb6 = i n Shumiak­
ina-Aseev, Rostov 1993.
After the text-move (7 ttJf3), a
good high-level example went 7 ... g6 8
ttJfd2 ! ttJbd7 9 ttJc3 a6 10 a4 b4 1 1
ttJce4 as 1 2 'fib3 i.e7 1 3 ttJc4 ttJxe4
14 i.xe4 ttJb6 I S ttJxb6 'fixb6 1 6 .ih6
j.a6 17 .id3 .if8 1 8 .ixf8 �xf8 1 9
0-0 �g7 20 e4 .l:!.he8 = Portisch-Kor­
chnoi, Lucerne OL 1 982.
C)
4 ttJf3
Now:
Cl: 4...exd5 17
C2: 4...d6 18
CI)
4...exd5 5 cxd5 d6 6 e4 (D)
B b) 9 . . . .ixf3 10 ttJxf3 g6 1 1 0-0
Relying upon the tactic 6... ttJxe4??
7 �a4+. 6 e4 is a clever move-order
and probably the most important of
these irregular lines. It will mainly
17
appeal to those who want to play .id3,
0-0, and perhaps h3, either foregoing
ttJc3 in favour of ttJd2-c4 or delaying
ttJc3 so as later to transpose to a fa­
vourable version of the Modem Main
Line (see Chapter 9).
6...g6
This natural move leads to some
murky complications, although it seems
ultimately quite playable for Black.
An alternative is 6... a6: 7 .id3 (7 a4
ttJxe4; 7 ttJc3 i.g4 8 �a4+ ttJbd7 9
ttJd2 bS 1 0 'fic2 ttJb6 =) 7 ... i.g4 8
ttJbd2 ttJbd19 h3 and now:
a) 9 ...ttJeS ! is a possibility which I
was afraid of when I played White in
one game; e.g., 1 0 .ie2 ttJxf3+ 1 1
ttJxf3 .ixf3 1 2 .ixf3 g6 with a posi­
tion in which the simplification pre­
vents White from overrunning Black's
position by f4 and eS, and allows the
g7-bishop and queenside majority to
exert themselves. I would nevertheless
be interested to see how this would
play out in practice, since the two
bishops are always a force.
.ig7 1 2 i.f4 'fie7 (perhaps better is
12 .. :irVc7 ! ? with the idea that 13 eS can
be met by 1 3 ...dxeS 1 4 ttJxeS ttJxdS I S
ttJxd7 ttJxf4 - Shaked) 1 3 'iid2 0-0 1 4
:tfe l 1:.fe8 I S 1:.ad l , Yermolinsky­
Shaked, USA Ch (Denver) 1 998, and
here Shaked gives the interesting line
IS ...ttJhS ! ? 16 .ih2 ( 1 6 .igS .if6 1 7
.ie3 .ig7 "with the idea . . . ttJeS" -
Shaked, but Black would have to be
careful not to rush; e.g., 1 8 b3 ttJeS ?
19 ttJxeS 'iixeS 20 g4 ttJf6 2 1 .if4 'fie7
22 eS must favour White) 1 6 ... ttJeS 1 7
.ie2 ttJf6 ! ? 1 8 ttJxeS ttJxe4! 1 9 ttJxg6
18 THE GAMBIT GUIDE TO THE MODERN BENONI
( 1 9 1Wc2 i.xe5 20 i.xe5 'ifxe5 2 1 i.d3
f5) 1 9 ...hxg6 20 'ifc2 lbf6 =.
7 i.d3
7 i.b5+ is apparently pointless, but
gets played in such positions: 7 ...i.d7
(or 7 ... lbbd7 ! =) 8 .txd7+ lbbxd7 9
0-0 i.g7 10 i.f4, Naer-Kostenko, Tula
1 999, and now one approach is just
10 ...0-0 1 1 .txd6 .:te8 12 e5 lbe4 ! ? 1 3
lbc3 lbxd6 1 4 exd6 i.xc3 1 5 bxc3
lbb6, which is at least equal. 1 O...'i'b8
1 1 lbc3 0-0 is also interesting, since
Black has ideas like ....:te8 and ... b5,
or, in response to a4, ...c4 and ...lbc5.
7....tg7
After 7 ... a6 8 h3 .tg7 9 0-0 b5 1 0 a4
b4 1 1 lbbd2 lbbd7 1 2 a5 White is sim­
ply better.
8 0-0 0-0 (D)
But not 8 ....tg4? 9 'ifa4+ lbbd7 1 0
e5 ! .
w by 1 2...liJb6 !, equalizing on the spot)
This little-played position is ignored
by theory. Here are some plausible
continuations:
a) 9 %:tel .tg4 10 lbbd2 lbbd7 1 1
h3 lbe5 ! 1 2 .tc2 lbxf3+ 1 3 lbxf3
.txf3 1 4 1Wxf3 lbd7 =.
b) 9 h3 b5 ! 10 .:tel ( 1 0 a4 c4 1 1
.te2 b4 1 2 .txc4 lbxe4 1 3 .:tel l:te8 =;
for example, 1 4 .tb5 .td7 15 lbbd2
lbc5 1 6 lbc4 1he l + 1 7 lbxel a6 1 8
.txd7 lbbxd7 ! with the point 1 9
lbxd6? lbe5 + ; 1 0 lbbd2 a6 1 1 a4 c4
1 2 .tc2 i.b7 =) 1 O. . .c4 1 1 .tc2 lba6
1 2 lbc3 i.d7 1 3 .tf4 b4 14 lbe2 l:te8
15 liJg3 'i'b6 with a very unclear posi­
tion that I would assess as dynami­
cally balanced. This is one of those
cases I mentioned above in which 1 6
e5? ! dxe5 1 7 lbxe5 exposes White's
d-pawn after 1 7 . . .%:tad8 or 1 7 . . ..tb5
intending ....:tad8.
c) 9 lbbd2 .tg4 10 a4 (after 10 h3
i.xf3 1 1 lbxf3 ':'e8 12 %:tel lbbd7 1 3
i.f4 c4 1 4 .txc4 liJxe4 1 5 'i'c2 f5 I
think that Black is just active enough)
1 O. . .lbbd7 (perhaps 1 O. . .lbfd7 intend­
ing . . .lbe5 is simpler, in view of 1 1
ttJc4 ttJe5 or 1 1 h3 ttJe5 1 2 i.e2 lbxf3+,
etc.) 1 1 lbc4 lbb6 (or here I l .. .lbe8,
intending . . .lbe5; then 1 2 .tf4 is met
1 2 lbe3 c4?! (last chance? Black still
had 12 ....txf3 1 3 1Wxf3 liJfd7 !) 1 3 .te2
( 1 3 .tc2 .txf3 1 4 1Wxf3 ':'c8) was
played in From-Vistisen, Danish Ch
1 989. I think that White is better now;
on 1 3 ....txf3, he plays 14 gxf3 ! ( 14
.txf3 .:te8 1 5 a5 lbbd7 16 liJxc4 lbxe4
17 .txe4 %:txe4 1 8 lbxd6 .:td4 1 9 'i'c2
'i'b8 !) 1 4...liJh5 1 5 a5 lbd7 1 6 lbxc4
'i'h4, and now something like 1 7 .te3.
All told, Black's play is satisfactory
in these 6 e4 lines, but he should be
aware that they exist.
e2)
4 d6!?
•.•
 INTRODUCTION AND EARLY MOVES
This very rare move is a clever way
of bypassing the line above, since now
S e4? is bad in view of S ... lDxe4, and S
lDc3 exdS 6 cxdS is a normal Benoni.
White can try to go his own way by S
dxe6 ! ? fxe6 (S ... i.xe6 ! ?) 6 lDc3 lDc6
with play similar to Line A (4 dxe6),
or by S lDc3 exdS 6 lDxd5 ! ? lDxdS
(6 . . . g6 ! ? 7 i.gS i.g7) 7 �xdS lDc6 8
lDgS 'fie7 (8 ...'fif6? 9 lDxh7 !) 9 i.d2
h6 10 lDf3 i.e6 1 1 'fie4 0-0-0 + with
...dS next, Galliamova-Yakovich, Nov­
gorod 1 997. Clearly Black has no
problems in either case.
0)
4 lDc3 exdS
Now:
Dl: S lDxdS 19
D2: S cxdS 20
01)
S lDxdS
As far as I can see, this move doesn't
even appear in ECO. One could not
call it a common continuation, but it
will appear from time to time, espe­
cially at lower levels.
S lDxdS 6 'fixdS (D)
•••
I have seen no serious examples of
6 cxdS d6 (e.g., 7 i.d2 .i.e7 8 .i.c3
i.f6). Such a position is easier for
Black to play than the main lines, an
example of the general tendency for
simplified positions to favour Black.
This theme will recur as the book pro­
ceeds.
6...lDc6
The following should also equalize:
a) 6...i.e7 discourages b3 in view
of.. .i.f6; e.g., 7 e3 lDa6 ! ? 8 Wid 1 0-0 9
19
B
.i.e2 lDc7 10 lDf3 dS = Binks-Trejos,
Istanbul OL 2000.
b) 6 ... d6 7 e4 lDc6 8 i.f4 i.e6 9
'fid2 'iff6 = O.Gonzalez-Anton, Spain
1 996.
c) 6 . . . lDa6 ! ? 7 i.d2 lDc7 8 Wid3 dS
9 cxd5 WixdS 10 'fixd5 lDxdS 1 1 e4
lDf6 12 0-0-0, Cabrera-Romero, Span­
ish Cht 1 990, and now 1 2 ...i.d7 (or
12 . . . a6) 1 3 i.c4 i.c6 looks equal.
7 lDf3 d6 8 e4
Or:
a) 8 lDgS 'fif6 9 i.d2 ! ? .i.e6 1 0
Wid3, Maciag-Porod, Recklinghausen
1 999. Now Black should grab the
pawn by 1 O ...'fixb2! 1 1 i.c3 'ifa3.
b) 8 g3 .i.e6 9 "iHe4 i.e7 10 .i.g2
0-0 1 1 0-0 dS = Mawunto-Myo, Ja­
karta 1 997.
8 i.e6 9 �dl i.e7 10 .i.d3 0-0
•.•
Or 1O ... lDeS =.
After the text-move ( 1 0...0-0), S.Nik­
olic-Rogers, Moscow GMA 1 989
continued 1 1 0-0 a6 12 i.f4 i.f6 1 3
�d2 Wie7 1 4 lDgS h6 ( l 4. . .lDeS I S
lDxe6 Wixe6 also leads to equality) I S
lDxe6 fxe6 1 6 .i.g3 .i.d4 with an equal
position.
20 THE GAMBIT GUIDE TO THE MODERN BENONI
02)
S cxdS d6 6 e4
6 h3 is a rare move-order:
a) 6... g6 7 e4 ! ? (7 ttJf3 is Line A of
Chapter 9; one idea after 7 .i.gS .i.g7 8
e3 h6 9 .i.h4 is 9 ...ttJa6 with the idea
...ttJc7, since h3 is wasted in that case)
7 ... .i.g7 8 ttJf3 transposes to Line B2
of Chapter 9.
b) 6 ...a6 side-steps that line: 7 a4
(7 ttJf3 bS 8 a4 b4 9 ttJbI .i.b7 1 0 .i.gS,
Flottat-Durret, Paris 1993, 1O ... h6 ! 1 1
.i.xf6 �xf6 is good for Black) 7 . . :fii e7
8 ttJf3 g6 (8 ... .i.fS ! ? is also a good op­
tion with this particular move-order)
transposes to Line A of Chapter 9.
6 g6 (D)
..•
This standard position normally
leads to the main lines, discussed in
later chapters, as noted below. The de­
viations are innocuous at best.
7 .i.bS+
An attempt to simplify, but in gen­
eral, White doesn't want early simpli­
fication in the Benoni. The argument
that White is exchanging off his bad
bishop ignores the fact that he is also
solving Black's difficulties by ex­
changing the problem bishop on c8 or
allowing it to develop more easily.
Also, while Black' s other bishop on
g7 is technically 'bad ' , White would
gladly trade it for his good one!
Transpositions and oddities:
a) 7 .i.gS is discussed in Chapter 3 .
b ) Irregular lines with 7 .i.d3 are
examined at the beginning of Chapter
7.
c) 7 ttJge2 .i.g7 8 ttJg3 is also ana­
lysed at the beginning of Chapter 7.
d) 7 h3 is found in the introduction
to Chapter 9.
e) 7 g3? ! (this doesn't go well with
e4) 7 ... .i.g7 8 .i.g2 0-0 9 ttJf3 ? ! (9
ttJge2 lle8 { intending 1 O...bS } 10 a4
ttJa6 1 1 0-0 c4) 9 ... bS ! 10 ttJd2 b4 1 1
ttJe2 l:te8 + I.Stein-Kundin, Israeli U-
1 8 Ch 1 999; Black has active pieces,
and d3 is a weakness.
f) After the conventional 7 .i.e2
.i.g7, there are a number of continua­
tions that don't quite fit into other
chapters:
fl) 8 h3 0-0 9 ttJf3 .l:!.e8 10 ttJd2 is
just a Classical Main Line with a dubi­
ous tempo spent on h3. The ... ttJa6-c7
plan should gain in strength.
f2) 8 h4 ! ? has been answered by
8 ...hS, which is fine, but it seems to me
that 8 ...'fie7 also deserves attention;
e.g., 9 'fic2 (9 hS? ttJxe4 10 'fia4+ .id7
+; 9 f3 0-0 10 .i.gS h6 1 1 .i.e3 ttJhS)
9 . . .h6 10 hS gS 1 1 .i.e3 ! ? 0-0 and
White has some problems developing.
f3) 8 g4 ! ? 0-0 9 h4 (9 gS ttJe8 10 h4
fS) 9 ...l:te8 10 f3 a6 (or 1O...ttJa6) 1 1 a4
ttJbd7 1 2 hS ttJeS 1 3 ttJh3 lIb8 14 ttJf2
bS ! I S axbS axbS 1 6 .i.xbS .i.d7 1 7
 INTRODUCTION AND EARLY MOVES
i.e2 l::tb4, Kharlamov-Kogan, Vilnius
1 966. Nunn says that Black has "ex­
cellent play for the pawn"; and in any
case, Black can play just about any
plan (such as 1 O...LUa6 and I l . ..LUc7),
as White has neglected to develop and
his kingside advance doesn't achieve
much.
f4) 8 i.e3 a6 9 a4 �e7 ! disturbs
White's development.
fS) 8 LUf3 (White aims for the
Classical Main Line of Chapter 1 0
while ensuring that he can meet . . ..Jtg4
lines withLUd2) 8 ...0-0 9 0-0 and now
9 . . J:te8 is the main line of Chapter 1 0,
but for those who want to avoid ...l:te8
systems, Black can also play 9 ... a6 1 0
a4 i.g4, when 1 1 i.f4 i.xf3 1 2 .Jtxf3
is Line B 1 1 2 of Chapter 9, and 1 1
LUd2 i.xe2 1 2 �xe2 is note 'b3 ' to
Black's 9th move in Line A of Chapter
1 0. Both variations are safe and rec­
ommended for Black.
7 LUbd7 (D)
•.•
All the reasonable choices are good
here, such as 7 ...LUfd7 and 7 ... i.d7 8
a4 i.g7 9LUf3 0-0 10 0-0 lie8 1 1 l:te l
a6 1 2 i.xd7 LUbxd7 1 3 i.f4 'Wic7 14
'iVb3 LUg4 I S llad l LUgeS = Usachyi­
Ophoff, German seniors Ch (Weil­
burg) 1 998.
S i.f4
To tie the f8-bishop to the d-pawn.
Instead, 8 a4 a6 9 i.xd7+LUxd7 is al­
ready equal, while trying for a Pawn­
Storm Attack by 8 f4 ! ? gives up the
bishop-pair for very little: 8 . . . a6 9
21
i.xd7+LUxd7 10LUo i.g7 1 1 0-0 0-0
1 2 a4 11b8 = Tosic-G.Kuzmin, Alushta
1994.
S LUh5 9 i.e3 i.g7 10 g4!? LUhf6
•.•
n g5?!
Too greedy.
n LUh5 12 i.e2 0-0 13 i.xh5
...
gxh5 14 �xh5 �a5
Black has a big advantage. Mtin­
der-Wockenfuss, West Germany 1 974
went IS .Jtd2 ( I SLUge2LUeS { intend­
ing ... i.g4 or ...LUd3+} 1 6 0-0-0 .Jtg4
1 7 'ii'h4 i.xe2 1 8LUxe2LUc4 wins for
Black) I S ...LUeS 1 6 �e2 c4 1 7 'litfl
i.g4 ! ? (or 17 ... fS) 18 f3 i.hS 1 9LUdi
'iVb6 20 i.c3 fS with an ideal attack.
These deviations from the standard
Modem Benoni should not concern
Black. Most of them either release
White's hold on the centre or neglect
development. The exception is 4LUf3
exdS S cxdS d6 6 e4, but Black has
plenty of play in that case as well.
2 Systems with �f4
1 d4 lbf6 2 c4 c5 3 d5 e6 4 lbc3 exd5 5
cxd5 d6
In this chapter, we look at a wide
variety of systems with i.f4 for White.
These are traditional variations, like
those of Chapters 3 and 4, which are
not in vogue and considered by some
to be worked out. But the i.f4 ap­
proach is underrated, in my opinion,
and can cause Black considerably
more trouble than other unfashionable
lines. In fact, if it weren't for the popu­
larity of the Modern Main Line in
Chapter 9 and the Taimanov Attack in
Chapter 5, I think that we would see
more of i.f4 at the top levels.
There is a certain logic to White's
strategy, since i.f4 directly targets d6,
the primary weakness in Black's posi­
tion. Once Black plays ...i..g 7, White
can often increase that pressure by the
manoeuvre 'i¥a4+ and if Black replies
... i.d7, then '1lVb3. Then the queen at­
tacks b7, and the bishop on f4 threat­
ens the d6-pawn. Another strategy for
White is simply to take direct aim at
d6 by, for example, playing lbf3-d2-
c4, perhaps with h3 at some point to
give the bishop an escape square (h2)
versus ... lbh5. Finally, of course, Black
has the general problem that a natural
move like ... lbbd7 leaves his d-pawn
en prise, and therefore he may need
to play a move like .. :fle7 or ... "fic7
before he might like to commit his
queen.
On the other hand, i..f4 is not par­
ticularly forceful and allows Black the
lUxury of early queenside expansion
without having to fret over e4, f4 and·
e5 . The bishop on f4 will also be ex­
posed to attack by . . .lbh5 and White
may have to take time to prepare
against that move.
In contrast to some Benoni varia­
tions, I don't believe that the estab­
lished theory on the i.f4 systems
always suffices to give Black an even
game. Thus I will sometimes propose
other ways to equalize, accompanied
by considerable analysis. When I by­
pass a well-established solution, I will
try to give a brief impression of why I
have done so.
Having said that, this chapter's ma­
terial is organized as follows:
A: 6 i.f4 22
B: 6 lbf3 with 7 i.f4
(and 7 'ii'a4+) 24
C: 6 e4 with 7 i.f4 37
As much as in any other chapter, the
lines here can easily transpose into one
another. It may be best to ignore ques­
tions of move-order until one gets a
general feel for the lines involved.
A)
6 i.f4 (D)
 SYSTEMS WITH i.f4
White doesn't normally play this
move so early. Both sides have various
transpositional possibilities, but there
are also independent continuations of
note, including the exciting main line
that follows.
6 g6
•.•
A true Benoni. Since Black hasn't
committed to this move yet, he has
some unusual alternatives:
a) 6 ... iLe7 keeps an eye on d6.
Uhlmann-Vasiukov, Gotha 1 9S7 con­
tinued 7 e3 (7 e4! 0-0 8 �e2 :e8 9
4:Jf3 iLf8 1 0 4:Jd2 looks more chal­
lenging; on 10 ... 4:Ja6, 1 1 0-0 tiJc7 1 2
a4 b6 1 3 :b 1 ;!; intending b4 at some
point is an idea, but Black's position is
solid) 7 . . .0-0 8 h3 'iVaS 9 4:Jf3 ! ? (9
'iWd2 a6 intending . . .bS) 9 . . .4:Je4 1 0
'iVc2 fS 1 1 iLd3 4:Ja6 ! , intending to
meet 1 2 iLxe4 by 1 2 ...4:Jb4 1 3 'iVbl
fxe4 1 4 ii'xe4 :f7 IS g4 (versus
...iLfS) I S . . .ii'a6 with an attack; e.g.,
16 'iitd2 ! ? c4 1 7 4:Jel �f6 18 a3 ii'aS
and the d-pawn hangs.
b) 6. . . a6 7 a4 and then:
bI) 7 ... 4:JhS looks odd, but makes
some sense; e.g., 8 �d2 g6 9 e4 iLg7,
23
and now Zagorskis-Kovacevic, Cap­
pelle la Grande 1 998 continued 1 0
iLe2 ( 10 g4 ! ? 4:Jf6 1 1 g S 4:Jfd7 1 2 h4
ii'e7 1 3 �c2 is equal but contains
risks for both sides) 1O ... tiJf6 1 1 4:Jf3
iLg4, and Black has a reasonable ver­
sion of a main line Classical Benoni
(Chapter 1 0), since White's extra move
�d2 gets in the way.
b2) 7 ... g6 is an obvious response,
trying to transpose into a 'normal'
Benoni while eliminating White's
�a4+ option. Then 8 tiJf3 �g7 9 e4
iLg4 (9 ...0-0 1 0 tiJd2 is dealt with in
note 'c' to White's 8th move in Line
B22) is another Classical Benoni, un­
less White plays 10 'fj'b3, whereupon
1 O... �xf3 1 1 ii'xb7 tiJbd7 1 2 gxf3
transposes to note 'c' to Black's 1 0th
move in Line B 1 1 3 of Chapter 9.
7 "a4+
This move is most consistent with
the themes of this chapter. Instead, 7
4:Jf3 and 7 e4 are dealt with in Lines
B2 and C, but the following continua­
tions are unique:
a) 7 'iid2 a6 8 a4 ii'e7 (8 ...�g7 9
iLh6 0-0 =) 9 'fj'e3 (9 �h6 is more in­
teresting) 9 .. :ir'xe3 10 iLxe3 4:Jbd7 =
Petronic- Kosanovic, Yugoslav Cht
(Cetinje) 1 993. Black has good devel­
opment and a sound structure for the
endgame.
b) 7 e3 a6 (7 ... �g7 8 4:Jf3 trans­
poses to note 'a' to White's 8th move in
Line B22) 8 4:Jf3 bS 9 a4 ! ? b4 10 tiJbl
�g7 1 1 iLc4, Zagorskis-Kleeschaet­
zky, Berlin 1 993, and now 1 1 ...4:JhS !
1 2 iLgS �aS looks good, hitting b2;
e.g., 1 3 'iVb3 h6 14 iLh4 gS IS �g3
4:Jxg3 and ...4:Jd7.
24 THE GAMBIT GUIDE TO THE MODERN BENONI
7....td7
In such positions, ...ttJbd7 ties Black
to the defence of d6, and is generally
undesirable.
8 �b3 (D)
B . . . ttJxdS .
This situation will become quite fa­
miliar in the sections that follow, al­
though the exact position before us
has rarely arisen in practice. Having
lured Black's bishop to d7, White
takes aim at b7.
8...bS!?
A n extremely common pawn sacri­
fice in the Benoni designed to convert
a relatively passive position into one
with a lead in development and open
lines. If White accepts the pawn, he
will lose influence in the centre and be
exposed to attack down the b-file.
While promising and entertaining,
8 ...bS isn't strictly necessary. Perhaps
Black didn't like the looks of 8 . . .'iVc7
9 e4 .tg7 10 ttJbS ( 1 0 ttJf3 is Line C,
which 8 ...bS side-steps) l O....txbS 1 1
.txbS+ ttJbd7, which gives White the
two bishops. But in fact, Black stands
well in that case. The e4-pawn hangs
and ...a6 followed by ...bS has to be
dealt with. Moreover, Black's knights
have good prospects in such positions,
as we will see in similar positions
throughout the book.
9 ttJxbS .i.xbS 10 'ii'xbS+ ttJbd7 11
iid3
Black threatened . . .1:tb8 as well as
11 ..:iNb6
Also of interest is 1 1 . ..ttJb6 1 2 e4
�e7 ! ? After 1 3 0-0-0 .tg7 1 4 f3 0-0
I S ttJe2, Black plays l S . . .ttJhS with
the idea of . . .fS.
12 b3
Black has good play after 1 2 'ii'e 3+
Wd8 ! or 1 2 .tel 'iWb4+ 1 3 .td2 'ii'xb2
14 .tc3 'ii'a3 =.
After 1 2 b3, the game Hausner-Bel­
aska, Prague 1 99 1 continued 1 2 ... .tg7
1 3 .i'1d1 0-0 14 ttJf3 c4 ! ? (an aggres­
sive move, but 14 .. JHe8 ! looks stron­
ger, planning 1 5 e3?? 'iWb4+ or I S
ttJd2 ttJh5) 1 5 bxc4 l:tab8 1 6 'iVc2
(better is 1 6 ttJd2 ! , although White is
still underdeveloped) 1 6...ttJc5 1 7 i.d2
1!i'b2 with considerable pressure for
Black.
B)
6 ttJf3 g6
This section deals with ttJf3 and
.tf4, delaying or foregoing e4. Of
course, 7 .tf4 is the main move here,
but the rare 7 'ika4+ also intends .tf4,
so we will look at it as well:
Bl: 7 �a4+!? 24
B2: 7 i..f4 26
B1)
7 'iWa4+!? (D)
 SYSTEMS WITH �f4
This check has been played re­
cently, drawing the attention of some
strong players. White's idea is to wait
a bit before committing to SLf4, and to
avoid early ... a6 lines like 7 iLf4 a6.
Notice that the similar 7 e4 iLg7 8
'iYa4+ is discussed in note 'c' to
White's 8th move in Line B2 of Chap­
ter 9.
After 7 iVa4+, I will concentrate on
the reply...
7...SLd7
Nevertheless, 7 . . .ttJbd7 seems quite
playable. A recent game went 8 iLf4
a6 9 e4 ttJh5 ! ? (the less committal
9 . . . .l:.b8 seems safer: 10 ..Ite2 b5 1 1
'iYc2 WIIe7 ! ? 1 2 ttJd2 ..Itg7 =) 1 0 ..Itg5
�e7 ! ? (a new idea) 1 1 ..Ith6 ( 1 1 ..Ite3
0-0 12 �c2 ttJg7 1 3 SLd3 f5) 1 l . ....Itf6
12 'iYc2 SLg7 1 3 SLxg7 ttJxg7 1 4 a4 0-0
1 5 iLe2 f5 1 6 exf5 ttJxf5 1 7 0-0 ttJf6 =
Cifuentes-Marin, Barcelona 2000.
8 'iYb3 'ilc7 9 e4
After 9 SLf4, 9 . . .SLg7 transposes to
Line B22, while 9 ...ttJh5 ! ? is a logical
option.
9....i.g7 10 h3
25
An original position which I be­
lieve is critical to the 7 ... SLd7 defence.
Instead, 10 ..Itf4 transposes to Line C.
10 0-0 (D)
•••
ll..lte2
This seems most logical, but an
analysis of possible future options is
in order, if only because this variation
might soon grow in popularity:
a) 1 1 SLf4 'ue8 1 2 ttJd2 ttJxe4 ! 1 3
ttJcxe4 f5 1 4 O-O-O ! ( 1 4 ..Itxd6 �xd6
1 5 �xb7 'iYb6 1 6 'iYxa8 fxe4 is very
good for Black; e.g., 1 7 ttJc4 'ilb4+ 1 8
Wd l ..Itxb2 ! -+) 1 4. . . fxe4 1 5 ttJc4 b5
16 ttJxd6 J:l.f8 17 SLg3 c4 1 8 �b4!? ( 1 8
'iYc2 �a5 1 9 Wbl ttJa6 with a strong
attack based upon . . . c3) 1 8 ... ttJa6 1 9
�a3 �b6 looks very promising for
Black; e.g., 20 WIIe3 WIIxe3+ 2 1 fxe3
..Ith6! or 20 ttJxe4 :ae8 2 1 f3?? b4.
b) 1 1 ..Itd3 l:Ie8 1 2 0-0 c4 ! with
equality.
c) 1 1 SLe3 .I:i.e8 1 2 ttJd2 ttJxe4 ! ?
( 1 2 ... a6 1 3 a4 ttJxe4 i s also possible)
1 3 ttJcxe4 ( 1 3 ttJdxe4 f5 1 4 ttJxc5
'iYxc5 15 SLe2 .uxe3 ! ? 1 6 fxe3 �xe3
00 ) 1 3 ...f5 (D) and White has:
26 THE GAMBIT GUIDE TO THE MODERN BENONI
w f3.l:Iab8 1 6 �a4 tDhS ! 17 i.bS tDb6
c l ) 14 tDxcs f4 ! ? I S tDxd7 ( I S
�xb7? �xcS) I S ... tDxd7 = 1 6 i.bS
fxe3 17 fxe3 tDcs 1 8 'tWa3 a6! 1 9 i.e2
( 1 9 i.xe8 i.xb2) 1 9 ...�e7 +.
c2) 1 4 tDxd6 ! ? 'tWxd6 IS �xb7 f4
1 6 tDc4 �a6 1 7 'ilixa8 fxe3 1 8 fxe3
tDc6 1 9 tDd2 ( 1 9 �xe8+ i.xe8 20
dxc6 i.xc6 is comfortable for Black)
1 9 . . . 'tWaS 20 �xe8+ i.xe8 2 1 dxc6
i.xb2 22 i.c4+ �g7 23 l:ldl �c3 with
at least enough counterplay for Black.
1l l:le8 12 tDd2
.••
Here Black seems well-poised for
the characteristic Benoni sacrifice:
12 bS!?
•••
If needed, a safer line would be
12 ...tDa6 1 3 0-0 :ab8 (or 1 3 ...:e7 1 4
l:lel l:lae8) 14 l:le l ( 1 4 a4 tDb4) 14. . .bS !
I S tDxbS �aS .
13 tDxbS
1 3 i.xbS tDxe4 14 tDcxe4 fS gives
counterplay on the dark squares and
down the b-file. Just for example: I S
f3 i.xbS ( 1 S ... i.d4 1 6 i.d3 i.c8 ! ?) 1 6
�xbS tDd7 1 7 0-0 fxe4 1 8 fxe4 ( 1 8
tDxe4 tDf6 1 9 tDxf6+ i.xf6) 1 8 . . .tDeS
1 9 'ili'e2 l:lab8 20 tDf3 tDxf3+ 2 1 gxf3
'iWb7 ! intending ...�xdS .
13...i.xbS I4 i.xbS
The alternative 14 �xbS tDbd7 I S
1 8 iVc2 l:le7 gives Black more than
enough for a pawn, with ...tDg3, ... fS,
and ... i.eS or ... i.d4 to come.
14 tDbd7 (D)
•.•
White's e-pawn still hangs and I S
f3 tDhS i s very difficult for him.
IS �a4 tDxe4! 16 tDxe4 .l:Ixe4+ 17
'iixe4 'tWaS+ 18 i.d2 'iixbS
Black has more than enough com­
pensation, threatening b2 and ...tDf6
or ...tDeS; e.g., 1 9 i.c3 tDf6 20 'iff3
(20 i.xf6 :e8 ! ) 20 . . ..:te8+ 2 1 �d1
tDe4 ! 22 i.xg7 �xg7 and Black wins!
The pawn sacrifice 12 ... bS is an in­
structive example of Benoni dyna­
mism; it would be fun to see more of
this variation in practice.
82)
7 i.f4
This is White's most popular move­
order, introducing a very dangerous
system which requires careful han­
dling. Once again, the main idea is
 SYSTEMS WITH iLf4 27

�a4+ followed by 'ib3. We will look b) 8 lDe4 ! ? lDxe4 9 'ii'a4+ lDd7 10

at two systems against 7 .i.f4, intro­ 'iWxe4+ .i.e7 ( 1 O ... �e7 1 1 'ii'xe7+ <i;;xe7
duced by these moves: 1 2 e4 .i.g7 1 3 lDd2 ! ? bS 1 4 0-0-0 lDeS
B2l: 7 ..a6. 27 =) 1 1 i.xd6 lDf6 1 2 'ii'xe7+ 'ii'xe7 1 3
B22: 7 ..i.g7
.. 33 .i.xe7 �xe7 i s fine for Black, as White
cannot hold on to the d-pawn; e.g., 1 4
82 1 ) ':'c 1 b6 I S lDeS lDxdS 1 6 g3 .i.e6 1 7
7 a6 (D)
.•. i.g2 f6 1 8 lDc4 bS 1 9 lDaS ! ? l:tac8 20
0-0 c4 2 1 ':'fd l ':'hd8 =+= Dreev-Ivan­
chuk, Lvov 1 987. 22... c3 is threatened,
and if 22 �fl , 22 . . . b4! looks good.
W
82 1 1 )
S lDd2 b5
Not 8 ...i.g7? 9 lDc4 0-0 1 0 a4!
(safest) 1O ... lDe8 1 1 �d2 and White
has an advantageous version of the
Knight's Tour Variation (Chapter 4);
nor 8 . . .lDh5? ! 9 'ilYa4+ ! 'iVd7 (9 ....i.d7
10 'ilYe4+ 'fie7 1 1 i.xd6) 1 0 'ii'e4+
.i.e7 1 1 i.h6 and Black is tied up; e.g.,
l l . ..bS 12 'fic2 fS ? ! , A.Petrosian­
This is a slightly safer move than Yudasin, Moscow 1 989, and now 1 3
7 ... i..g7, keeping d6 guarded for a mo­ e4! ±, according to Petrosian.
ment and discouraging 'ii'a4+. Black 9 a4 b4
pays a price in development, but the 9 ... lDhS 10 i.e3 b4 1 1 lDce4 lDd7
counterattack by ...bS is his justifica­ 1 2 lDc4 lDdf6 is also fine.
tion. White has three replies to 7 ... a6: 10 lDce4 (D)
B2ll: S lDd2 27
B212: S a4 28
B213: S e4 30

Otherwise:
a) 8 'iWa4+ ! ? is untried, but 8 ...b5 9
lDxbS .i.d7 10 e3 'ii'b6 1 1 .i.gS lDg8
1 2 'iWe4+ .i.e7 1 3 lDa3 'ii'xb2 1 4 lDc2
'iWc3+ looks fine for Black. Less clear,
but perhaps also satisfactory for Black,
is 8 . . . .i.d7 9 'ii'c2 (9 �b3 bS) 9 ...'ii'c7
10 e4 ( 1 0 a4 .i.g7 1 1 h3 ! ? resembles
later lines) 1 O... lDhS 1 1 .i.e3 .i.g7.
28 THE GAMBIT GUIDE TO THE MODERN BENONI
10 i.b7!
.•.
Or:
a) 10 ...tZJxdS? I I i.xd6 ! i.xd6 (if
1 1 ...tZJe3, then 12 �b3 ! tZJxn 13 �dS !)
1 2 tZJc4 i.e7 13 tZJed6+ i.xd6 14
..wxdS ±.
b) 1 O...tZJhS 1 1 i.gS f6 12 i.e3 fS
1 3 g4! fxg4 ( 1 3 ... f4 1 4 i.xcS !) 14 tZJc4
is probably fine for Black, but it's very
messy.
l l i.g5
1 1 e3 is dangerous if Black grabs on
dS , but simply 1 1 . . .tZJxe4 1 2 tZJxe4
i.e7 is fine; e.g., 1 3 i.c4 ( 1 3 i.h6 fS
14 tZJg3 i.f6 =) 1 3 ... fS I 4 tZJd2 i.f6 1 S
�c2 0-0 intending . . .'Wie7 and . . .tZJd7.
1l ... i.e7 12 i.xf6 ii.xf6 13 tZJc4
i.e7
The game is equal, Litinskaya-Prud­
nikova, USSR worn Ch 1 986.
82 1 2)
8 a4 i.g7 (D)
Black can play 8 ...'Wie7 ! ? to prevent
9 e4 of the next note, but there's no
need to do so.
w tal; for example, 1 0 'ilVb3 ! ? ( 1 0 tZJd2
9 h3
This restrained move may be the
most difficult line for Black to meet.
White preserves his bishop against
... tZJhS and plays for the simple e3 and
tZJd2-c4. Others:
a) 9 e4 and now:
a l ) After 9 . . .0-0 ! ?, 1 0 i.e2 i.g4
could follow, when 1 1 0-0 i.xf3 1 2
i.xf3 transposes to Line B 1 1 2 of
Chapter 9. 10 tZJd2 is treated very
briefly in note 'c2' to White's 8th move
in Line B22.
a2) However, much more ambi­
tious (and probably better) is 9 ... i.g4!
10 'Wib3 ( 1 0 i.e2 i.xf3 1 1 i.xf3 0-0
1 2 0-0 transposes to Line B 1 1 2 of
Chapter 9) 1O ...i.xf3 1 1 'i¥xb7 tZJbd7
1 2 gxf3, when we have transposed to
note 'c' to Black's 1 0th move in Line
B 1 1 3 of Chapter 9, which is perfectly
satisfactory for Black.
b) 9 tZJd2 tZJhS 1 0 i.e3 fS 1 1 tZJc4
( 1 1 g3 tZJd7 1 2 tZJc4 tZJeS =) 1 l .. .�c7
(to meet 1 2 as with 1 2 ...tZJd7) is safe
and equal.
c) 9 e3 and now:
c 1 ) 9 ... 0-0 10 h3 transposes to the
main line.
c2) 9 ... i.fS ! ? would be experimen­
tZJhS +) 1 O...'Wic7 ! ? 1 1 i.xd6 'ilVxd6 1 2
�xb7 0-0 1 3 'ilVxa8 tZJe4 intending to
meet 14 tZJdl ( 1 4 ZIc 1 'iYb6) by 1 4 ...c4
I S i.e2 llVb4+ 1 6 <it>n i.xb2 1 7 .l:.a2
c3 with good chances.
c3) 9 ... i.g4! ? 10 'YWb3 ( 1 0 i.e2 0-0
{ 1 O . . . i.xf3 1 1 i.xf3 O-O } 1 1 tZJd2
i.xe2 12 �xe2 tZJhS !) 1O ...i.xf3 1 1
gxf3 ! ? ( 1 1 'iVxb7 tZJbd7 1 2 gxf3 .l:.b8
1 3 'iWxa6 l:txb2 is critical, as in line
'a2' ; this would be a bit more solid for
 SYSTEMS WITH i.f4
White due to the trade-off e3 for e4)
1 l . . .'Yi'c7 1 2 i.xd6 'ifxd6 1 3 'iWxb7
ttJfd7 ! ? 14 a5 0-0 1 5 'iixa8 i.xc3+ 1 6
bxc3 'iVf6 17 l:tc1 'iVxf3 1 8 l:tg 1 ttJe5 1 9
i.e2 'iVe4, Gunawan-Hulak, Sarajevo
1988 . According to Psakhis, Black has
a "strong attack".
In any case, both 9 ... 0-0 or 9 ...i.f5
are good options after 9 e3.
9 0-0
.•.
Now 9 ...i.f5 is pointless in view of
10 ttJd2 ! .
10 e3 (D)
Or 1 0 ttJd2 ttJe8 1 1 ttJc4 lbd7 ! 1 2
i.xd6 ( 1 2 lbxd6?? lbxd6 1 3 i.xd6
i.xc3+ 14 bxc3 'iff6 -+) 1 2...lbxd6 1 3
lbxd6 'iVb6 14 lbc4 'iVxb2 ! 1 5 lbxb2
i.xc3+ 1 6 'iVd2 i.xd2+ 1 7 �xd2 lbf6
=t .
B
10...lbh5
Direct. Another possible course is
1O .. :�e7 1 1 .ie2 ( 1 1 lbd2 lbh5 1 2
i.h2 lbd7 1 3 lbc4 lbe5 00) 1 l ...lbbd7
( 1 1 .. .'ud8 ! ? 1 2 lbd2 lbe8 1 3 lbc4 lbd7
14 0-0 b6) 1 2 lbd2 l:tb8 1 3 0-0 lbe8 14
i.h2 f5 = Deshmucke-Marin, Calcutta
1999.
29
11 i.g5 (D)
This is considered best, because 1 1
i.h2 f5 intends to cut off the bishop
and to attack e3 with ...f4: 12 i.e2 ( 1 2
'ilkd2 i.h6 ! ? { or 1 2 .. .'i!i'e7, now that
lbd2-c4 is no problem } 1 3 i.d3 f4 14
0-0 lbd7 1 5 l:tael :b8 1 6 'ilkc2 �g7 =
Genov-Pigusov, Berlin 1 992) 1 2 ... f4
1 3 e4 lbd7 14 0-0. This can be met by
14 . . .�h8 and ... J:.b8, as in Furman­
Forintos, Oberhausen Echt 196 1 , or by
14 . . .'ti'e7; e.g., 1 5 lbd2 lbe5 1 6 i.xh5
gxh5 17 'iVxh5 i.d7 ( 1 7 ... f3 ! ? imme­
diately also deserves consideration)
with compensation, according to Bar­
lov. Indeed, White is hard-pressed for
a move; e.g., 1 8 a5 f3 ( 1 8 . . J1f6 ! ?) 1 9
i.xe5 ( 1 9 lbxf3 i.e8 ! ) 1 9 ...fxg2 20
�xg2 .ixe5 2 1 l:tg1 �h8 and Black is
for choice.
B
11 'iVb6!?
.•.
This move is untried and unsug­
gested, but may save the day. Other­
wise:
a) 1 l . ..f6? ! 12 i.h4 g5 13 lbd2 is
good for White, I.Sokolov-Ki.Geor­
giev, Groningen 1 994.
30 THE GAMBIT GUIDE TO THE MODERN BENONI
b) 1 1 . .:iVc7? ! 1 2 .i.e2 ( 1 2 liJd2 ! ?)
1 2 .. .f5 1 3 liJd2 liJf6 14 liJc4 (Barlov)
also favours White.
c) As is the case with the text­
move, 1 1 . .:iVa5 ! ? is untried. Then 1 2
'iWc2 i s met by 1 2 . . J�e8 1 3 liJd2 .i.f5 ! ,
but 1 2 'iVd2 gains a tempo with the
threat of 1 3 .i.e7, though 1 2 . . . .:te8 1 3
g4 liJf6 may be alright anyway.
12 "iWc2
Leaving room for liJd2-c4. After 1 2
.:tbl , 1 2. . .h6 1 3 .i.h4 g 5 14 liJd2 .i.f5 !
1 5 e4 .i.g6 is equal.
12 h6! 13 g4
•.•
Or:
a) 1 3 a5 �c7 14 g4 liJg3 ! 1 5 fxg3
hxg5 16 liJxg5 �e7 is at least equal.
b) 1 3 .i.h4 ! ? .i.f5 ! ( 1 3 . . . g5 14 g4 is
unclear) 14 e4?! ( 14 'iVd2 liJf6 15 a5
�c7 =, intending ...liJe4) 14 .. J�e8 1 5
liJd2 'ili'b4 1 6 .i.d3 liJf4 + . This would
not have worked with the bishop on
g5 .
13 liJg3! 14 fxg3 hxg5 15 liJxg5
•..
.i.h6
Or 1 5 ...'iVd8 ! ? with the idea ... f5.
16 liJge4
1 6 h4 .i.xg4 1 7 a5 �d8 1 8 liJe6 ! ?
'iWe7 1 9 liJxf8 'iVxe3+, and while 20
liJe2 holds on, I'd rather be Black.
16 .i.xe3 17 liJf6+ rJ;;g7 18 a5
.••
'ili'd8 19 liJce4 .i.d4 20 g5 .i.f5
Black's activity and bishops out­
weigh the temporary bind.
8213)
8 e4
Allowing Black to play ...b5 in order
to undermine his central pawn-struc­
ture.
8...b5 (D)
If this seems too risky, 8 .. :ife7 has a
solid reputation: 9 .i.e2 (9 'iVe2 liJbd7
1 0 0-0-0 liJg4 !?; 9 e5 liJbd7 is given
by Kapengut, when 1 0 �e2 dxe5 1 1
liJxe5 liJh5 1 2 d6 "iVe6 looks fine for
Black; 9 liJd2 b5 1 0 .i.d3 .i.g7 1 1 0-0
0-0 is dynamically equal) 9 . . . liJbd7
(9 . . . liJxe4?? 10 'iWa4+) 10 0-0 .i.g7
(safer than 1O ...b5 ! ? 1 1 .i.xb5 ! ? axb5
1 2 liJxb5, although that is unclear) 1 1
liJd2 ( 1 1 a4 0-0 1 2 liJd2 b6 ! ? or
1 2 ... liJe5) l l . ..b5 ("unclear" - Kapen­
gut). Now if White plays f3, ... liJh5 is
effective, so Black has serious pres­
sure on the e-pawn. Probably White
should again play 1 2 .i.xb5 ! axb5 1 3
liJxb5 .:ta6 1 4 liJc7+ rJ;;f8 1 5 liJxa6
.i.xa6 16 .:tel liJh5 17 .i.e3 .i.xb2 1 8
.ubI .i.e5 1 9 'iVa4 i.c8 20 .i.h6+ (20
liJc4 �g7) 20....i.g7 21 .i.e3 .i.e5 with
a repetition.
w
9 'iWe2
Other moves are considered satis­
factory for Black:
a) The position after 9 liJd2 .i.g7
10 .i.e2 0-0 can arise via a number of
move-orders. 1 1 0-0 liJe8 1 2 .i.g3
 SYSTEMS WITH i.,f4
(versus ... f5, but Black plays it any­
way; 1 2 'flic2 f5 1 3 h3 g5 ! ? 14 i.h2 f4
= Anastasian-Moldobaev, Belgorod
1989) 1 2 . . .f5 1 3 exf5 i.xf5 14 Si.g4
tL'ld7 = Safin-Gelfand, USSR jr Cht
(Kramatorsk) 1 989.
b) 9 �c2 i.g7 10 i.e2 ( 1 0 tL'ld2
0-0 1 1 i.e2 l:te8 1 2 0-0 l:.a7 1 3 i.f3
i.g4 14 i.xg4 tL'lxg4 1 5 h3 tL'le5 = For­
intos-Paavilainen, Tallinn 1 986; 1 0
a4?! b4 1 1 tL'ld l 0-0 1 2 i.e2 'flie7 1 3
tL'ld2 ':e8 1 4 0-0 tL'lxe4 1 5 tL'lc4, Kal­
entarian-Moldobaev, Blagoveshchensk
1988, and now 1 5 ... i.b7 1 6 i.f3 "iVf6
or even 1 5 ... g5 looks strong) 10 . . .0-0
1 1 0-0 :e8 1 2 a3 ! ? ( 1 2 tL'ld2 b4 1 3
tL'la4 tL'lxd5 1 4 exd5 ':xe2 1 5 �d3 ! is
given as favouring White by Hebert,
but 1 5 . . . .:xd2 ! 1 6 �xd2 i.d7 can
hardly be bad) 1 2 . . .b4 ( l 2 . . . .:a7, with
the idea . . . .:ae7, and 1 2 . . . i.g4 are
sound alternatives) 1 3 axb4 cxb4 1 4
tL'lb5 l:1xe4 1 5 i.g3 b3 ! 1 6 �d3 i.f8 1 7
tL'lfd4 i.b7 i s unclear, Abarca Aguirre­
Klinger, Kiljava jr Wch 1 984.
c) 9 i.d3 i.g4 ! ? (9 ... i.g7) 10 h3
i.xf3 1 1 'flixf3 is solid for Black, as
usual in these ...i.g4 and ...i.xf3 lines:
1 l . . .i.g7 (interesting is 1 l . ..tL'lbd7 1 2
0-0 tL'le5 ! ?) 1 2 0-0 0-0 1 3 ':fe 1 tL'le8 ! ?
1 4 a4 b4 1 5 tL'ld l tL'ld7 1 6 �e2 ( 1 6 :te2
i.d4 ! 1 7 'iVg3? ! tL'le5 += Kallai-Foisor,
Val Thorens 1 987) 1 6... a5 17 i.c4 l:::ta7
is perhaps slightly better for White,
Liogky-Levin, Nikolaev 1 987. Black
would like to get moves like . . .tL'lb6
and . . .l:tae7 in.
d) 9 e5 dxe5 10 tL'lxe5 i.d6 1 1 i.e2
0-0 1 2 0-0 �c7 ( l 2 ... l:::te8 1 3 tL'lc6
tL'lxc6 1 4 i.xd6 tL'ld4 = Kapengut) 1 3
i.f3 J:.e8 ( 1 3 ...b4 ! ? 14 tL'la4 tL'lbd7) 1 4
31
tL'lg4 i.xg4 1 5 i.xd6 �xd6 1 6 i.xg4
and now 16 . . .tL'lbd7 1 7 i.xd7 �xd7 1 8
�f3 r3i;g7 was equal in Zielinski-Jaw­
orski, Bielsko-Biala 1990, but 16 ....:a7 !
(to capture on d7 with the rook, or to
double) looks quite good; e.g., 1 7 �f3
( 1 7 i.f3 tL'lbd7 1 8 l:.e l tL'le5 1 9 lIc l
l:::tae7) 17 . . .b4 1 8 tL'la4 ':ae7 1 9 l:.ac l
':e4 ! 20 i.h3 c4 and the d-pawn will
fall.
9 i.e7 (D)
...
Not 9 ... i.g7?? 1 0 i.xd6 ! 'iVxd6 1 1
e5 �e7 1 2 d6, etc.
W
A truly remarkable position, which
illustrates the resilience of the Benoni.
When White first realized that he
could force this position, he must have
thought: 'That's the end of 7 . . . a6 ! ' .
White has a large lead in development,
the prospect of 0-0-0, and deadly­
looking attacking ideas involving e5
and if . . .dxe5, d6. Black's bishop is on
the 'wrong' square e7, and White even
has notions of i.h6 at some point, pre­
venting Black from castling! But in
fact, White seldom even enters into
this position any more, which is a
32 THE GAMBIT GUIDE
tribute to the defensive powers con­
ferred by Black' s pawn-structure.
10 �c2
A rather tame move, but perhaps
best. Having provoked Black's bishop
to e7 instead of g7, White loses a
tempo in order to develop. Instead, 10
ii.h6? ! is useless in view of 1 O. . .lLlg4,
but the alternatives produce some very
entertaining play:
a) 10 0-0-0 0-0 1 1 h3 ( 1 1 eS lLlg4 !
1 2 lLle4 dxeS 1 3 lLlxeS lLlxeS 14 ii.xeS
lLld7 I S i.f4 lle8 = Alburt-D.Gure­
vich, USA Ch (Estes Park) 1 986)
1 1 ...�e8 12 'iVc2 ii.f8 and Black looks
fine; e.g., l 3 eS ? ! dxeS 14 lLlxeS??
ii.d6 0- 1 Ra.Garcia-Psakhis, Andorra
1 997, since I S lLld3 c4 is winning for
Black.
b) 10 eS dxeS 1 1 i.xeS lLlbd7 and
now:
b 1 ) 1 2 d6 lLlxeS l 3 dxe7 lLlxf3+ +.
b2) 1 2 i.d6 Wf8 + Kapengut. The
point is that White's d-pawn becomes
weak.
b3) 1 2 0-0-0 lLlxeS l 3 lLlxeS 'iVd6
14 lLlc6 �f8 (this is likely +, due to the
weak d-pawn) I S g3 ( 1 S h3 'iVf4+ 1 6
.l:i.d2 ii.d6 1 7 g 3 �fS 1 8 �e3 �g7 and
Black connects his rooks, Lundin-Ilie,
Lugano 1 985; IS lLlxe7 �xe7 16 d6
'iVxe2 17 ii.xe2 ii.b7 1 8 ii.f3 ii.xf3 1 9
gxB �g7 + Milovanovie-Hulak, Pula
1 990) l S . . . i..fS ( 1 S . . . ii.b7 is also
played) 1 6 ii.g2 �e8 1 7 l:i.he1 ii.d8 ! 1 8
�d2 �g7 1 9 l:i.xe8 ':xe8 20 lLlxd8
.l:txd8 21 h3 hS 22 'iVgS ne8 23 'iYd2
b4 24 lLle2 lLle4 with a strong attack,
H.Pedersen-Emms, Esbjerg 1 996.
b4) 12 ii.g3 0-0 1 3 d6 :'e8 14 dxe7
l:txe7 l S lLleS (D) and now:
TO THE MODERN BENONI
b4 1 ) This widely-quoted line from
Ilie's analysis concludes with the
move l S . . .b4' ! ' . However, after 1 6
lLle4 ! , White seems to have a great
game. Since 16 ...lLlxe4 1 7 lLlc6! is bad
for Black, he probably has to try
1 6 . . . l:i.xeS 1 7 ii.xeS lLlxeS 1 8 .l:!.d 1
'iVe7, when after 1 9 lLld6 i.g4 20 B
ii.hS (20 ...l::i.d8 2 1 lLlc4) 2 1 lLlc4 l:te8
22 lLlxeS 'iYc7 23 Wf2 the attack is fin­
ished.
b42) So l S . . . lLlxeS ! must be cor­
rect; e.g., 1 6 i.xeS b4 1 7 :td1 ( 1 7 lLldl
lLld7 { 1 7 ... lLldS ! ? } 18 f4 f6 19 "iVB
l:ta7 20 i.c4+ �h8 is messy, but is
probably better for Black) 1 7 .. :tliaS
( 1 7 ...'iVe8 ! ?) 1 8 .txf6 ( 1 8 lLldS lLlxdS
1 9 J:!.xdS ii.b7 +) 1 8 ... l:txe2+ 1 9 lLlxe2
ii.fS ( 1 9 . . . .tb7 ! ? 20 h4! "iVxa2 2 1 hS
�e6 22 .tgS is unclear) 20 lLlg3 (20
lLlc I ? b3+ 2 1 ii.c3 bxa2 22 ii.xaS a1 'iY
23 .tc3 .tc2 -+) 20..."iVxa2 2 1 ii.e2
(21 lLlxfS ? ! gxfS ! 22 f4 l:i.e8+ 23 �f2
c4 24 .te2 c3 +) 2 l . . .'iVe6 22 ii.gS
ii.c2 23 .l:i.d2 b3 intending . . . 'iYeS and
...'iVxb2 and you have to like Black's
position.
10...0-0 (D)
 SYSTEMS WITH iL.f4
W Now the black rook is trapped, but
11 iLe2
1 1 a4 b4 1 2 tDbl ( 1 2 tDdl b3 ! 1 3
'iVc4 tDbd7 gives Black the initiative,
Hear-Renet, Hastings 1 987/8) 12 .. J�e8
( l 2 ... tDhS 1 3 iLh6 �e8 yielded satis­
factory positions in two of Wedberg's
games as Black against Yakovich; still
more aggressive is 1 2 ...b3 ! ? 1 3 'iVxb3
tDxe4 14 iLd3 tDf6 I S 0-0 tDbd7 1 6
tDc3 l:tb8 1 7 'it'a2 �e8 = Damaso-de
Firmian, Lisbon 2000) 1 3 iLd3 iLf8
14 0-0, Yakovich-Totsky, St Peters­
burg 1 994, and now Psakhis recomm­
ends simply 1 4...tDbd7 =, but 14 ...b3
may also suffice.
11 .l:teS
..•
1 1 ...iLg4 is also sufficient for equal-
ity.
12 tDd2
To stop ... tDhS and ...iLg4.
12 iLfS 13 0-0 b4!?
•.•
13 ... tDbd7 14 l':.fel .l:i.b8 = was
Flear-Schulte, Oakham 1 988, which
saw the interesting strategy I S b3 tDeS
16 a4 b4 1 7 tDd 1 tDhS !? 1 8 iLxhS gxhS
19 tDb2 tDg6 1/2- 1/2. Black has active
play; e.g., 20 iLe3 iLg7 2 1 l'tabl h4 22
h3 iLc3 23 tDbc4 'it'f6 24 �edl tDf4.
33
14 tDdl tDxd5 15 exd5 ':xe2 16
iLe3
Black will get two active bishops and
a pawn for the exchange: 1 6 ... as 1 7
'it'd3 iLa6 1 8 tDc4 'iVh4 1 9 l:t c 1 lhe3
20 tDdxe3 tDd7 2 1 'it'd l fS 22 �e l tDf6
23 'iVf3 tDe4 24 tDxaS ! ? iLg7 2S tDec4
iLd4 ! and now, instead of 26 l:tc2??
iLxc4 27 tDxc4 tDxf2 ! winning for
Black, S.Schneider-de Firmian, Co­
penhagen 1 999, White has to try 26 g3
tDgS ! 27 'iVg2 (27 'iVe2 'iVh6 28 �f1
tDh3+ 29 �g2 tDxf2 +) 27 ...'iWh5 28
�h l , when 28 ...iLxc4 29 tDxc4 l:txa2
30 tDxd6 �xb2 3 1 �e8+ �g7 32 .l:i.e7+
�h6 ! ? (32 ... �f8 =) maintains a few
chances for Black, although White
should reach full equality.
822)
7 iLg7 (D)
•••
W
This is our second, more ambitious,
repertoire choice. I believe that 7 ... iLg7
is perfectly sound, but it is probably
more difficult to play than 7 ... a6 and
requires a lot of understanding.
34 THE GAMBIT GUIDE TO THE MODERN BENONI
8 'iVa4+
Easily the most important move.
Notice that S e4 0-0 9 4Jd2 transposes
to note 'c' . If White doesn't play S
'iVa4+, the move 7 ...J.. g7 shows its vir­
tues by saving a tempo and being less
weakening than 7 ... a6:
a) S e3 (not even mentioned in
ECO) S ... O-O 9 h3 4JeS (easiest) 1 0
.Jte2 4Jd7 I I 0-0 4Je5 1 2 .Jtxe5 ( 1 2
4Jd2 f5 ! = ; 1 2 'iVc2 4Jxf3+ 1 3 .Jtxf3 a6
14 a4 rLbS intending ...b5, followed by
. . . .Jtd7 if necessary) 1 2 . . . dxe5 (this
pawn-structure tends to favour White
if ... a6 and a4 are thrown in, but here it
is very solid) 1 3 4Jd2, and now
1 3 ...4Jd6 (the ideal blockader) 1 4 'iVb3
b6 = seems more accurate than 1 3 ...f5
14 'iVb3, Tal-Spassky, USSR jr Ch
(Leningrad) 1 954.
b) S h3 0-0 9 4Jd2 4JeS 10 4Jc4 is
almost the same as the note to White's
1 0th move in Line B 2 1 2: 1O ... 4Jd7 ! 1 1
.Jtxd6 ( 1 1 4Jxd6?? 4Jxd6 1 2 .Jtxd6
.Jtxc3+ 1 3 bxc3 'iVf6 -+) 1 l .. .4Jxd6
12 4Jxd6 iVb6 1 3 4Jc4 'iVxb2! 14 4Jxb2
J..xc3+ 1 5 'iVd2 J..xd2+ 1 6 �xd2 4Jf6
=F.
c) S 4Jd2 0-0 (simpler than S ... 4Jh5
9 'iVa4+ ..tfS 1 0 .Jte3 a6, which is dy­
namically equal) 9 e4 (a relatively
harmless position which can be ar­
rived at with a variety of move-orders;
9 4Jc4 is the Knight's Tour Variation
of Chapter 4) and then:
c l ) 9 ...4Jg4 1 0 .Jte2 ( 1 0 4Jc4? J..d4!
1 1 J..g3 f5 !) 1O . . .4Je5 1 1 0-0 f5 is one
way to equalize.
c2) 9 ... a6 1 0 a4 and now 1 O ...4Jh5
1 1 J..e 3 f5 ! ? ( 1 l .. .4Jd7 1 2 .Jte2 4Je5 ! ?
1 3 0-0 'iVh4) 1 2 exf5 .Jtxf5 1 3 g4
i.d4! is wild and unclear, Strang­
miiller-Krausner, Germany 1 992. For
those with attacking inclinations, I
would direct your attention to 1O ...4Jg4
(as in line ' c l ' ) intending to meet 1 1
4Jc4? with 1 l .. .J.. d4 ! - this idea is
playable with or without the inclusion
of ... a6 and a4.
c3) 9 ...4JeS 1 0 J..e2 f5 1 1 exf5 (else
... g5 and ...f4) 1 l .. .J..xf5 1 2 .Jtg3 4Jd7
1 3 0-0 4Je5 1 4 4Jde4 a6 1 5 a4 :bS
16 :e l 4Jc7 17 a5 b5 = Gavrikov-de
Firmian, Biel 1 995. Then I S axb6
l1xb6 would be a typical example of
pressure down the b-file.
8 J.. d7 9 'iVb3
.••
9 'iVc2 0-0 1 0 e4 ( 1 0 J.. xd6? J..f5
1 1 e4 l:.eS =+= 1 2 0-0-0 J..h6+ 1 3 4Jd2
.Jtxe4 !) 1 O ...'iVe7 1 1 J..e2 :eS 1 2 4Jd2
b5 ! 1 3 0-0 a6 ( 1 3 ...b4 14 4Jb5 4Jxe4 1 5
:ae l !) 1 4 J..f3? ! ( 1 4 1He l b4 1 5 4Jd l
i.b5 = ) 14 ...b4 + Herzog-Klinger, Zug
1 985. Black intends ... J..b5 followed
by ...4Jbd7.
9 'iVc7
•••
This is the older, less recommended
move, but I think it holds the balance.
By contrast, after the popular sacri­
fice 9 ...b5 (which Kapengut calls the
"modem alternative"), I think that 1 0
i.xd6 ! i s favourable to White ( 1 0
4Jxb5 i.xb5 1 1 'iVxb5+ 4Jbd7 1 2
i.xd6 4Je4 1 3 .Jte5 0-0 1 4 J..xg7
..txg7 15 'iVa4 :bS ! ! 16 'iVxe4 l1xb2 is
an amazing sacrifice which is still
holding up well). The crucial line is
1O ... 'iVb6 ( 1 O . . .c4 1 1 'iVd l 'iVb6 1 2
J..e5 b4 1 3 4Jbl ! ± ECO) 1 1 J..e5 0-0
1 2 e3 c4 1 3 'iVd l ( 1 3 iVb4 ! ? goes un­
mentioned) 1 3 ...4Ja6 (D) ' ! ' , accord­
ing to all Benoni sources, who agree
 SYSTEMS WITH i..f4
that the alternatives for Black are
weaker.
Here I think the major theoreticians
are much too kind to Black, giving
him full equality. The only sceptical
voice is David Norwood (normally
optimistic for Black), who comments:
"Although some Benoni players are
very happy to play this variation, I
have a deep suspicion that it is not en­
tirely sound for Black. White should
be able to keep the pawn and emerge
with a reasonable position." I agree,
and thus cannot recommend 9 . . .bS .
For those interested, some places to
begin looking are 14 l:Ie l , to answer
...b4 with tLlbl , 14 ..te2 (underrated),
and even the main line: 14 �d4 l':tfc8 ' ! '
I S .:te l b4 1 6 tLlbl ! (previously 1 6
tLld 1 was played), a sample l ine being
16 ...�xd4 17 tLlxd4 tLlxdS 1 8 ..txg7
rJ;xg7 19 ..txc4 tLlb6 20 tLld2 tLlcS 2 1
0-0 tLlxc4 2 2 tLlxc4 tLld3 2 3 l':tc2 as 24
:Ld2 lhc4 2S l:txd3 and White has a
clear advantage.
We now return to the position after
9 .. :Wic7 (D):
35
10 h3
The main move is 1 0 e4, which we
will take up in the next section (Line
C) via 6 e4 g6 7 ..tf4 ..tg7 8 1lVa4+, etc.
The only other serious alternative is 1 0
tLld2 tLlhS, and now:
a) Once again, 1 1 ..txd6 'it'xd6 1 2
�xb7 i s messy. I think Black should
be OK after something like 1 2 . . .llVb6
1 3 llVxa8 0-0 14 tLlc4 �b4 I S l':tc 1
�xc4 1 6 �xa7 ..txc3+ 17 ':xc3 1lVxdS
1 8 e3 l':te8 1 9 l:txcs 'iVe4 20 .tIc 1 tLlf4
with the initiative.
b) 1 1 ..tgS (the book move) l l . ..h6
12 ..th4 gS 13 ..tg3 tLlxg3 14 hxg3 a6
(or 14 . . .0-0) I S a4 ..tfS 16 tLlc4 tLld7
1 7 as 0-0 1 8 tLla4 �ae8 ! with equal
play, Sturua-Eolian, Erevan 1 982. This
is an example of the common phe­
nomenon in which a knight on b6 will
be stranded away from the action.
10 h3 is a deceptive move which is
supposed to be somewhat better for
White. Fortunately, since White plays
rather slowly with moves like llVa4-b3,
h3 and e3, Black has time to develop
rapidly:
10...0-0 11 e3 tLla6
36 THE GAMBIT GUIDE TO THE MODERN BENONI
Black develops as quickly as possi­
ble, additionally clearing his first rank.
12 lDd2 :ab8
The move which has been played,
but it does cost a tempo. I think a good
alternative which illustrates Black's
counterplay is 12 ...:fd8 1 3 lDc4 lDe8
14 a4 ..tf5 ! ? (D).
For example: 1 5 g4?! ( 1 5 ..te2 'iile7
1 6 lDd2 g5 1 7 ..th2 lDf6 1 8 0-0 lDb4
=) 1 5 ... ..txc3+ ! ( 1 5 ... ..tc8 is equal) 1 6
bxc3 ( 1 6 �xc3 ..te4 1 7 ':'gl ..txd5)
16 .....te4 1 7 ':'gl ..txd5 1 8 ':'d1 ..tf3 =t.
13 a4
1 3 lDc4 lDe8 1 4 lDe4 ( 1 4 a4 lDb4
transposes to the main line) 1 4...b5 ! 1 5
lDcxd6 'ii'a5+ 1 6 lDc3 l:Ib6 1 7 lDxe8
l:txe8 and Black has all kinds of play
involving ...b4 and ... ..ta4 or ...c4.
13...lDb4 (D)
14 lDc4
ECO gives 14 ..te2 as slightly better
for White, but 1 4 ...lDe8 1ooks satisfac­
tory: 1 5 0-0 ( 1 5 lDb5 ..txb5 1 6 axb5
lDf6 and now 1 7 lDc4 ':'fd8 =, or 1 7
..tf3 ':'fd8 1 8 0-0 a6 1 9 :tfc 1 �e7 20
b6 lDd7 2 1 lDc4 i.e5 ! =; 1 5 lDc4
w
..txc3+ ! ? and 1 6 ... lDxd5) 1 5 ... a6 1 6 a5
( 1 6 lDce4 ! ? b5 1 7 axb5 ..txb5 1 8
..txb5 ':'xb5 1 9 l:txa6 ! ? 'iilb7 20 .l:!.xd6
f5 ; 1 6 .l:f.fc l b6 1 7 lDce4 'iiid8 1 8 lDxd6
lDxd6 1 9 ..txd6 i.xa4 20 .l:f.xa4 'ii'xd6
is equal) 16 . . .b6 ! 1 7 axb6 .l:!.xb6 1 8
lDc4 :b8 with dynamically balanced
play.
14 lDe8 15 ':'dl a6 16 as i.f5
•.•
I am suggesting this because I don't
like 16 ... i.b5 ? ! 17 lDxb5 axb5 1 8
lDb6 .l:f.a8 ! 1 9 lDxa8 'iVxa5, Velichko­
Peresypkin, USSR 1 984. This is given
as unclear in the books, and in fact
Black won the game, but this is hard to
believe. For example, Bangiev gives
20 :d2 g5 ( ' ! ' ; 20...c4 2 1 ..txc4) 2 1
..txg5 lDf6, yet 22 lDc7 'fIxc7 2 3 i.xf6
..txf6 24 ..txb5 �a5 25 ..te2 leaves
Black short.
17 e4
Black keeps things level after 1 7
.l:f.d2 �e7 ! 1 8 ..te2 ltd8 1 9 0-0 g5 20
..th2 lDf6; e.g., 2 1 g4 ..tg6 22 f4 gxf4
(or 22 ... lDe4) 23 i.xf4 lDe4 =.
17..:iVe7 18 lDxd6?!
Better is 1 8 f3 ..td7 1 9 ..te2 ..tb5 20
0-0 f5 with equality, in view of 2 1
 SYSTEMS WITH i.f4
lDxbS axbS 22 lDd2 and now 22... CiJc7
or 22 ...gS and .. .f4.
18 CiJxd6 19 i.xd6 'i!i'xd6 20 exf5
..•
b5!
Introducing a typical trade-off of
a-file for b-file.
21 axb6
Giving up a piece by 2 1 fxg6 c4 22
gxh7+ �h8 23 i.xc4 bxc4 24 "iixc4
offers White no relief after 24 ...nfc8
2S �g4 .i.xc3+ 26 bxc3 �eS+ 27 'iir>f 1
lDxdS .
21...lbb6
Also reasonable is 2 1 ... �xb6 22
"iia4 (22 i.c4 CiJd3+! 23 �xd3 "iixb3
24 i.xb3 l1xb3) 22 ...�fe8+ 23 i.e2
i.d4 24 fxg6 hxg6.
22 CiJa4 :e8+ 23 i.e2 l:.b7 24 fxg6
White's problem is that he can't get
castled; 24 l:!.d2 nbe7 leads to similar
play.
24...1:tbe7 25 gxf7+ 'iir>xf7 26 l:%.d2
i.h6! 27 f4 i.xf4 28 0-0 'iir>g7
White has serious problems.
Apparently Black maintains the
balance after 1 0 h3. Aside from the
main line above, his alternative on
move 1 2 should be taken seriously.
To conclude, both 7 ...a6 and 7 ...i.g7
are satisfactory answers to 6 CiJf3 g6 7
i.f4, which is White's most popular
i.f4 system. 7 . . . a6 is probably the
safer choice, in that it prevents the
'i!i'a4+ manoeuvre and, incidentally,
avoids the play that follows in the next
section. Nevertheless, 7 . . . i.g7 may
create a type of tactical environment
that appeals more to one's taste.
C)
6 e4 g6 7 i.f4
37
Here White delays CiJf3 for a while,
and thus avoids some of Black's early
...i.g4 ideas.
7...i.g7
It may be possible to play 7 ... a6 and
try to transpose to Line B 1 of Chapter
9, but I'm not sure about the unique
position after 8 �a4+ ! ? i.d7 9 'i!i'c2 ! ?,
since ... a6, ... i.d7 and ... bS do not go
together so well here. Still, this is not
completely clear and worth investigat­
ing. There could follow 9 . . .'i!i'e7
(9 ...bS 1 0 CiJf3 "iie7 1 1 0-0-0 b4 1 2
CiJbl t) and then:
a) 1 0 CiJf3 i.g7 ( 1 0...bS 1 1 eS dxeS
{ I l .. .CiJhS 1 2 CiJe4 i.fS 1 3 'i!i'c3 } 1 2
0-0-0 exf4 1 3 .:tel i.e6 1 4 CiJgS i s very
messy, but ultimately seems better for
White) 1 1 0-0-0 0-0 1 2 eS dxeS 1 3
CiJxeS CiJhS 1 4 CiJxd7 CiJxd7 I S i.e3 bS
1 6 i.e2 CiJhf6 1 7 h4 ! ? t.
b) 10 O-O-O ! ? CiJhS 1 1 i.e3 i.g7 1 2
i.e2 CiJf6 ( 1 2. . .0-0? 1 3 i.xhS gxhS 1 4
CiJf3 fS I S i.gS �f7 1 6 l:%.hel ± ) 1 3
CiJf3 t.
8 �a4+
Or:
a) After 8 CiJf3 0-0, 9 CiJd2 trans­
poses to note 'c' to White's 8th move
in Line B22, while 9 i.e2 a6 1 0 a4
i.g4 1 1 0-0 i.xf3 1 2 i.xf3 transposes
to Line B 1 1 2 of Chapter 9, and 9 h3?
is note 'a' to White's 9th move in Line
B2 of Chapter 9.
b) 8 i.bS+ i.d7 9 i.e2 should be
met by 9 ... 'i!i'e7 1 0 CiJf3 ( 1 0 �c2 CiJa6
Hebert) 1 0...0-0 ( l 0...CiJxe4! ? 1 1 CiJxe4
1lYxe4 1 2 i.xd6 'i!i'b4+! 1 3 'iir>f 1 'i!i'b6
could be tried) 1 1 0-0 i.g4 (here
I l .. .CiJxe4 1 2 CiJxe4 �xe4 1 3 i.xd6
i.a4! is equal) 1 2 CiJd2 i.xe2 1 3 'i!i'xe2
38 THE GAMBIT GUIDE TO THE MODERN BENONI
ttJh5 1 4 i.e3 ttJd7 = Spassov-Ermen­
kov, Albena 1 975.
8 i.d7 9 �b3
...
Now 9 �c2 "fie7 puts immediate
pressure on the e-pawn (by contrast,
see the note to White's 9th move in
Line B22, i.e. 6 ttJf3 g6 7 i.f4 i.g7 8
"fia4+ i.d7 9 �c2).
9 �c7 10 ttJf3
••.
Not 1 0 i.xd6? �xd6 1 1 'i!Vxb7 'i!Vb6
1 2 "fixa8 �xb2 +.
The text-move brings us to an im­
portant position that could have arisen
via 1 0 e4 in Line B22.
10 0-0 (D)
•••
11 ttJd2!
This probably poses the most prob­
lems. Others are intriguing, but only
equal:
a) 1 1 i.d3 a6 ( 1 1 . . .ttJh5 1 2 i.e3
ttJa6 ! ? would compare interestingly
with the 1 1 ttJd2 main line) 12 a4 ttJh5
1 3 i.e3 i.g4 ! ? 1 4 ttJd2 ttJd7 1 5 h3 ( 1 5
f3 ttJe5 1 6 i.e2 i.d7; 1 5 0-0 ttJe5 1 6
i.bl c4 1 7 'ili'a3 i.d7 1 8 h 3 f5 ! + Kap­
engut; an exaggeration, but Black has
enough play) 1 5 ...ttJe5 1 6 i.f1 i.d7
1 7 g4 ( 1 7 f4 ttJg3) 1 7 ... f5 ! (just in
time, and justifying the position of the
knights on h5 and e5) 1 8 gxh5 ( 1 8 f4
ttJg3 1 9 fxe5 ttJxh 1 20 e6 f4 is 'un­
clear' according to Kapengut; in fact,
2 1 i.g l f3 22 ttJd l i.e8 probably fa­
vours Black) 1 8 . . .f4 1 9 hxg6 ( 1 9
i.xc5 'i!Vxc5 20 hxg6 hxg6 looks un­
clear, with dark-square pressure for
the pawn) 1 9 ... hxg6 20 0-0-0 fxe3 2 1
fxe3 b5 ! Kapengut. Black's knight i s a
giant on e5 and his queenside attack is
menacing.
b) 1 1 h3 l:te8 ( 1 1 .. .b5 ! ? Kuligow­
ski; then 1 2 i.xb5 i.xb5 1 3 "fixb5 is
critical) 1 2 ttJd2 ttJxe4 1 3 ttJcxe4 f5
with active play; e.g., 1 4 0-0-0 ( 1 4
i.xd6 �xd6 1 5 'i!Vxb7 fxe4 16 ttJc4 �f4
1 7 i.e2 { 17 'i!Vxa8 1:.f8 ! } 1 7 . . .i.d4!
with the idea that 1 8 0-0 is met by
1 8 ... i.xh3) 14 ... fxe4 15 ttJc4 b5 1 6
ttJxd6 1:.f8 1 7 i.g3 c4 1 8 "fic2 ii'a5 1 9
�bl ttJa6 + ; e.g., 20 a3 ttJc5.
c) 1 1 i.e2 gives Black a broad
choice of moves. Now the books offer
two moves that are theoretically ade­
quate. 1 1 ...b5 has been successful, but
it also requires lengthy and tactically
complex analysis. Since 1 1 i.e2 isn't
seen much, I'd rather recommend the
somewhat easier 1 1 ... ttJh5 1 2 i.e3
i.g4 (D) (it's worth mentioning that
1 2 ... a6 has also been successful, meet­
ing 1 3 ttJd2 with 1 3 ...b5 ; in that case
1 3 a4 allows an improved version of
the text after 1 3 ... i.g4) and now:
c l ) 1 3 h3 i.xf3 1 4 i.xf3 ttJd7 ! ? 1 5
i.xh5 gxh5 i s the typical trade-off of
black dynamism for a weakened
pawn-structure. Now 16 O-O ? ! is
frowned upon by theory due to Black's
 SYSTEMS WITH �f4
w B
activity following 16 ... f5 !, and 16 'iWd l
f5 17 'iWxh5 f4 1 8 .td2 liJe5 1 9 0-0
:f6 is also undesirable. So in Tim­
man-Ljubojevic, Amsterdam 1 972,
White continued 16 'iWc2 :ae8 1 7 0-0
a6 1 8 a4 'iWd8 ( 1 8 ... liJe5 ! ?) 1 9 a5 ! ?
'i¥o>h8 20 .l:.a4 'iWc8 2 1 liJd1 (Kapengut
gives 2 1 'iWe2 f5 22 'iWxh5 f4 with com­
pensation) 2 1 ...b5 22 axb6 liJxb6 23
1::ta2 f5 =.
c2) 13 0-0 ltJd7 1 4 h3 jtxf3 1 5
i.xf3 liJhf6 ( 1 5 ... a6 1 6 jtxh5 gxh5 is
similar to line 'c1 ') 16 a4 a6 1 7 a5 ( 17
jte2 .l:.fc8 ! ? intending ...c4 - Kapen­
gut; 17 .l:.fe 1 .l:.fe8 1 8 jtf4 b6 is modest
but sensible for Black, contemplating
a combination of ... liJe5, ...c4, ...:ab8
and ... b5) 1 7 .. J�fb8 1 8 .l:.a2 b5 1 9 axb6
l:!xb6 = Salov-Psakhis, Moscow 1 986.
The game continued 20 'iWc2 liJe8 2 1
i.e2 'iWd8 (intending . . .liJc7-b5) 22 f4
'VJIie7 23 .tf3 l:tab8 24 i.f2 liJc7 25
lie I , and here Psakhis likes 25 ... ltJb5
==
d) After 1 1 e5 (D) there has been
some strange analysis:
d 1) For example, after 1 1 .. .:e8,
held to be equal, everyone follows
39
Ubilava-Basin, Tbilisi 1 983, but after
1 2 0-0-0 liJh5 1 3 exd6 'iWa5 1 4 jtd2
that game continued 14 ... liJa6, al­
though White has simply 1 5 'iWxb7 !
and where's the compensation? If
15 ... liJb4, White can successfully cap­
ture on d7, but he also has 1 6 .tc4 ! .
Kapengut suggests 1 4...'iWb4, but 1 5
�xb4 cxb4 1 6 liJb5 looks very strong
( 1 6 ...ltJa6 1 7 ltJc7 !).
d2) 1 1 . ..liJh5 could be adequate,
but again, there's a flaw in the main
line after 1 2 exd6 ( 1 2 .te3 ! ? dxe5 1 3
0-0-0 jtg4 ! ? appears safe enough)
12 ...l:te8+ 1 3 jte3 ( 1 3 .te2?? .txc3+
14 'iVxc3 lIxe2+) 1 3 . . .'iWa5 and now:
d2 1 ) After 14 .te2 ! ? b5 ! 1 5 0-0,
the commonly cited line, Kapengut's
15 ...c4, meeting 16 'iWc2 with 16 ...b4
intending . . ..tf5, has the huge hole 1 6
jtxc4 ! bxc4 1 7 'iWb7, which i s even
worse than line 'd1 ' . So Black should
play 15 ... b4! 1 6 ltJd 1 ltJf4 17 jtxf4
( 1 7 i.c4 .ta4) 17 .. .lhe2 1 8 l:te1 ( 1 8
'iWc4 'iWb5 ! i s unclear) 1 8 . . J:he 1 + 1 9
liJxe1 .ta4 20 'iWc4 liJd7 with ideas
such as ... .tb5 and ... lIe8, achieving
equality.
40 THE GAMBIT GUIDE TO THE MODERN BENONI
d22) Although 1 4 tbd2 f5 looks
active for Black, one wonders about
1 5 g4 ! ? f4 1 6 tbc4. This may cast
1 1 . . .tbh5 into doubt.
d3) l l . ..dxe5 (I like this best) 1 2
.txe5 �e8 ( l 2 . . :iVb6 i s also fine, in
view of 1 3 'iVxb6 axb6 14 .tc4 .tg4!
or 13 .tc4 'iVxb3 14 axb3 { 14 .txb3
.tg4 = } 14 . . ..tg4 1 5 0-0 .txf3 1 6
gxf3 tbbd7 1 7 .td6 �fc8 with equal­
ity, Gahwens-Kaspret, Cologne 1 980)
1 3 0-0-0 'iVb6 (or 1 3 . . .'iVc8 - Kap­
engut) 14 'iVxb6 axb6 1 5 �bl , and
now 1 5 . . . .tf5+ is equal. A more dou­
ble-edged try would be 1 5 ... .tg4 16
.tg3 tbbd7; for example, 1 7 .td3 tbh5
1 8 tbb5 tbxg3 1 9 hxg3 tbe5, which is
unclear.
1l tbbS
•.•
Kapengut (in NIC Yearbook 18)
gives 1 1 ..JXe8 as playable. However,
in his main line, I don't like 1 2 .te2
tba6 1 3 0-0 l:.ac8 14 �fe l c4 ! ? 1 5
tbxc4 tbc5 1 6 'iVa3 .tf8 1 7 �xa7,
when White is just better.
12 .te3 (D)
12 tba6
..•
I am recommending this move,
which is barely known and not even
considered by Psakhis, Schneider or
some other sources. On an elementary
level, it develops a piece, connects
rooks, and creates no weakness. In ad­
dition, Black now has ways of improv­
ing his position, such as . . .�ab8 and
. . . b5, . . .f5, or even . . ..td4 ! ?
a ) Note that playing 1 2 . . ..td4? im­
mediately fails, but not to the book
suggestion 1 3 tbb5 ' ! ' .txb5 14 .txb5,
when in my opinion Black has equal
play with moves like 14 . . . tbd7 (or
14 . . ..txe3 ! ? 15 'iVxe3 f5 16 0-0 f4) 1 5
i.xd4 cxd4 1 6 'iVa4 tbc5 1 7 �xd4
'ika5. Rather, 1 3 i.xd4 ! cxd4 14 tbe2
tba6 1 5 tbxd4 is strong, when Black
lacks compensation; e.g., 1 5 . . .'iVa5
( l 5 . . .tbc5 16 �c3) 1 6 �xb7 ! tbc5 1 7
b4 , etc. But, significantly, one can see
that the inclusion of ... tba6 would
change this assessment. Compare what
follows.
b) The book alternative, which is
supposed to be equal for Black, is the
much-played and analysed 1 2 . . .f5 1 3
exf5 gxf5 , which is so established that
I should provide some indication of my
dissatisfaction: 14 i.e2 .te8 (the ad­
vance 14 . . . f4? has proven insufficient
in both theory and practice after 1 5
.txc5 ! , when one try i s 1 5 .. .f3 ! ? 1 6
.txf3 �xc5 17 tbde4 !) 15 tbf3 (threat­
ening tbg5) 1 5 . . .h6 (I have looked in
detail at alternatives, but don't believe
in them; for example, 1 5 ... f4 16 .td2
h6 17 0-0 tbd7 1 8 l:.fel a6 19 a4 �b8
20 'iVd I ! tbdf6 2 1 tbh4 or here 2 1
'iVc2; also falling short i s 1 5 . . .'iVe7 1 6
0-0 tbd7 17 l:.fe l f4 1 8 .td2 tbe5 1 9
 SYSTEMS WITH iL.f4
'iVd l ! fif7 20 liJe4 !) 1 6 0-0 a6 1 7 a4
ttJd7 (D).
W In looking for a better method for
This is a key position which Kap­
engut gives as leading to a variety of
'unclear' assessments. However, I be­
lieve that Black's kingside weaknesses
are very serious, and that a better as­
sessment is probably '±' . This is im­
portant enough to take a brief look at:
b 1 ) 1 8 l:tae 1 ':'b8 1 9 'iid 1 'iid 8,
when it seems to me that just 20 �c 1 !
is strong, protecting b2, so that ... b5
cannot be played. The weakness off5,
e6 and g6 will tell.
b2) Perhaps an even better idea is
18 l:tfe 1 l:tb8 1 9 'iid 1 (that move
again !) 1 9 ...'ii'd8 ( l 9 ...liJhf6 20 'iic2
ttJg4 2 1 �f4 liJge5 22 liJh4 !) 20 �c l !
(Kapengut's 20 g3? b5 ! is unclear)
20 . . . 'ii'f6 2 1 a5 �f7 22 ':'a4 l:tbe8
(22 ... l:tfe8 23 ':'h4 b5 24 axb6 ':'xb6 is
Kapengut's improvement, which he
considers "unclear", but White is
much better after 25 l:ta4! liJb8 26
liJd2) 23 l:th4 (23 'iib3 ;t) 23 ...:e7 24
g3 l:tfe8 25 �fl ':'xe l 26 liJxel 'iig6
27 �e2 liJhf6 28 liJg2 and White has
41
control of the key squares, Wegner­
A.Schneider, Hamburg 1 987. This is a
good illustration of Black' s difficul­
ties.
Thus I think 1 2 ...f5 is rather weak.
Black, I discovered that the text-move,
1 2 ...liJa6 (D), to which we now return,
has hitherto unnoticed virtues.
13 �e2
This is the only move given by Kap­
engut. But there are some untried al­
ternatives, including a very critical
one:
a) 1 3 �xa6 is almost always harm­
less or bad in such positions, since it
sacrifices light squares and the b-file:
1 3 . . .bxa6 14 0-0 f5 15 exf5 l:tab8
( 1 5 ...�xf5 =) 16 'iia3 gxf5 17 liJc4
�b5 (or 1 7 ...f4 1 8 �d2 f3) 1 8 liJxb5
axb5 1 9 liJa5 b4 20 'iia4 �xb2 and
one likes Black.
b) 1 3 f4? ! exposes White down the
e-file: 1 3 . . .l:tae8 ( l 3 . . .'itth8 14 �e2 f5
is also OK) 1 4 i.e2 �xc3 ! ? 1 5 bxc3
( 1 5 'ili'xc3 f5) 1 5 ...liJf6 ! ( l 5 ...�f5 is
just slightly better for Black) 16 �f3
42 THE GAMBIT GUIDE TO THE MODERN BENONI
( 1 6 eS dxeS I 7 0-0 e4 !) 1 6. ..lt:Jxe4 ! 1 7
lDxe4 fS +.
c) 1 3 lDc4 can hardly be more nat­
ural. I think that Black must respond
aggressively by 1 3 .. .1He8 ! 1 4 .te2
( 14 f3 fS ; 14 a4 can be answered by
14 ....txc3+ ! ? or 14 ...lDb4 I S i.e2 fS),
and now 14 ...bS ! is an attractive sacri­
fice: I S lDxbS i.xbS 1 6 1\¥xbS lDb4
1 7 1\¥a4 ! ( 1 7 lDxd6? 1\¥xd6 1 8 i.xcs
lDc2+! 1 9 �d2 { 19 'iti>n 1\¥f4 } 19 ...1\¥eS
-+; 1 7 0-0 1:.ab8 1 8 1\¥aS { 1 8 1\¥a4
':xe4 } 1 8 . . .1\¥ xaS 1 9 lDxaS lDf6 +)
17 ...1:.xe4 1 8 a3 lDxdS 19 i.f3 lDxe3 !
20 i.xe4 (20 fxe3 1:.ae8 and now 2 1
0-0-0 l:t4e6 22 ':xd6 i.xb2+! or 2 1
i.xe4 ':xe4 22 0-0-0 { 22 1:.d I ? 1\¥e7 + }
22... i.xb2+! with the point 23 �xb2
dS !) 20 . . .lDxc4 2 1 .txa8 1\¥e7+ 22
'i¥;>n lDd2+ 23 'iti>gl i.xb2 24 1\¥d l (24
':d l ? 'ife2) 24... .txal 2S 1\¥xd2 i.d4
+. Wild stuff!
13 1:.ab8 (D)
•..
1 3 ... .td4 ! ? still seems wrong in
view of 14 i.xhS .txe3 I S fxe3 gxhS
1 6 0-0 ;to
w ( 1 6 i.f4 lDe8 1 7 0-0 'iith 8 and ...fS)
14 a4
The obvious alternative is 1 4 i.xhS
gxhS, and now:
a) I S 1\¥d 1 fS 1 6 exfS .txfS 1 7
1\¥xhs lDb4 with the apparently forced
sequence 1 8 0-0 .td3 1 9 1:.fd 1 i.g6 20
1\¥gS l:tfS ! (20 ...lDc2 ! ? 2 1 l:.ac l 1:.fS =)
2 1 1\¥g3 .teS 22 1\¥h3 (22 f4 .txc3 23
bxc3 lDxdS) 22. . .lDc2 23 ':ac l lDxe3
24 fxe3 l:thS ! 2S 'ife6+ �h8 + with the
point that 26 g3?? fails to 26. . .i.fS.
b) I S lDe2 fS 1 6 lDf4 (16 i.f4 fxe4
1 7 lDxe4 1:.be8! gives Black an at­
tack), and an ambitious try is 1 6 ...'ifaS
( 1 6... fxe4 1 7 lDxhS .tfS ! is unclear),
hoping for 1 7 lDxhS? ! ( 1 7 a4 bS ; 1 7
0-0 i.bS ! ? 1 8 l:tfc l 1\¥b4 1 9 lDc4
i.xc4 20 1:.xc4 1\¥xb3 2 1 axb3 fxe4 22
1:.xe4 ':fS !) 17 ....ta4! 1 8 1\¥a3 ( 1 8 1\¥c4
.tbS 1 9 1\¥b3 lDb4) 1 8 ... lDb4 1 9 0-0
lDc2, when Black has the advantage:
20 1\¥d3 (20 lDc4? lDxa3 2 1 ltJxaS
.txb2) 20 ... lDxal 21 l:txal (21 lDxg7
.tbS) 2 l .. ..txb2 22 lDc4 (22 l:.bl ? !
"iic3 ! +) 22 ...1\¥c3 favours Black.
14 lDb4 15 .txh5
.•.
Logical. The other path is I S lDc4
lDf6 (now that the knight is ensconced
on b4, Black can regroup to prepare
for moves like ...bS and ... fS) 1 6 0-0
1 6...b6 17 l:.ac l a6 1 8 .tf4 lDe8 1 9
-Ufel 'ito>h8 ( 1 9. . .bS ! ? 20 axbS axbS 2 1
lDa3 ! 'ifaS ! 22 l:ta1 c4 23 1\¥dl 1\¥b6 is
obscure) 20 .td2 fS 2 1 lDe3 f4 22 lDc2
(22 lDfl bS 23 axbS axbS 24 lDd l
lDa6) 22 ...lDxc2 23 1:.xc2 l:ta8 =, with
the idea that 24 1\¥c4 is answered by
24 ...lDf6 ! ' In these positions, ... i.eS
can often be a useful preparation for a
kingside pawn advance.
15...gxh5 (D)
 SYSTEMS WITH i.,f4
16 ttJb5
What else? 1 6 0-0 f5 1 7 f3 a6 is
comfortable for Black; e.g., 1 8 a5 f4
19 .i.f2 ttJd3.
After the text-move ( 1 6 ttJb5), play
might go 16 ....i.xb5 17 axb5 f5 with
43
plenty of counterplay; for example, 1 8
f3 ! ( 1 8 exf5 l:txf5 hits the d-pawn)
1 8 .. .f4 ( 1 8 . . . a6 ! ?) 19 .i.f2 .i.xb2 ! 20
1ha7 lta8 21 J:txa8 �xa8 22 0-0 'ilig7
=, since 23 ttJc4? .i.d4 is suddenly
very bad for White.
The .i.f4 systems are not seen much
these days, and yet one can see how
difficult they are for both sides to play.
Careful study of this chapter can teach
one a lot about Benoni dynamism. Af­
ter mastering the tactical and strategic
concepts, it is probably worthwhile to
go back and examine move-order is­
sues. Finally, the reader may have no­
ticed how much of this chapter is bare
analysis, unsubstantiated by games. I
strongly encourage readers to do inde­
pendent work on the variations herein.
3 Systems with �g5
In this chapter, we look at an ambi­
tious complex of systems for White in­
volving the move .ig5. White wants
to force Black into a decision about
the pin on his f6-knight, hoping that
moves like ...h6 and ...g5 will prove
weakening. By activating the queen's
bishop, he also makes the move e3 at­
tractive, since that would otherwise
hem in the bishop. On the other hand,
this strategy often costs White the
bishop-pair by the sequence ' I ' ..h6
. C : 6 e4 g6 7 liJf3 i.g7 8 i.e2
'2' .ih4 g5 '3' .ig3 liJh5. White is
also more prone to allowing conces­
sions on the queenside, in part because
the dark squares on that wing are less
well guarded.
These systems have long been out
of fashion, and therefore quite a few of
the examples will be from older games.
Nevertheless, there are unexplored ar­
eas in any Benoni variation, and some
new ideas are also needed to uphold
older assessments. This slightly unset­
tled state of theory could be taken to
indicate that some of the .ig5 lines
will eventually regain their popularity.
In any case, the Benoni player would
be well advised to know the ideas be­
hind these variations and how to re­
spond to them.
1 d4 liJf6 2 c4 c5 3 d5 e6 4 liJc3
exd5 5 cxd5 d6
The material is divided into three
major categories, according to how
White proceeds. He can play: liJf3 and
.ig5 without e4; or liJf3 and .ig5 with
e4; or, finally, the combination of liJf3,
.ig5, e4 and .ie2 (irregular move­
orders will be discussed as we go
along).
Therefore:
A: 6 liJf3 g6 7 i.g5 44
B: 6 e4 g6 7 liJf3 .ig7 8 i.g5 50
0-0 9 .ig5 54
Needless to say, these systems can
often transpose to one another. Line B
contains a note on the move-order 7
i.g5, which often comes to the thing
after 7 . . ..ig7 8 liJf3, but can give rise
to some independent possibilities.
A)
6 liJf3 g6 7 Sl.g5
This is the most important move­
order these days, as systems with e4
are out of favour. Still, Black has noth­
ing to fear if he is prepared.
7 h6 (D)
.•.
A move designed to see where
White's bishop is going.
8 i.h4
The bishop maintains its pin, but
cannot now retreat to defend White's
queenside. 8 i.f4 is a poor version of a
.if4 system because the bishop can be
 SYSTEMS WITH Ji.g5
attacked immediately: 8 . . .lZ:lhS (or
perhaps 8 ... gS 1 ? 9 .i.c l { 9 .i.g3 lZ:lhS }
9... .i.g7 1 0 h4 g4 1 1 lZ:ld2 lZ:lbd7 1 2
lZ:lc4 lZ:lb6 =) 9 .i.g3 ( 9 .i.e3 1 ? .i.g7 1 0
'iVd2 lZ:la6 1 1 g3 lZ:lc7 1 2 a4 b 6 1 3 .i.g2
..tb7 !? 14 0-0 lZ:lf6 I S lZ:le l lZ:lg4 =)
9 . . . .i.g7 1 0 lZ:ld2 lZ:lxg3 1 1 hxg3 lZ:ld7
1 2 e3 0-0 = Filip-Rajkovic, Smeder­
evska Palanka 1 978.
Now there is another split:
AI: 8 g5 ..• 4S
A2: 8 .i.g7... 49
AI)
8. g5
..
Our main system, which is direct
and forcing.
9 .i.g3 lZ:lh5 (D)
10 e3
Or:
a) 10 'iWa4+ lZ:ld7 1 ? ( 1 0 .. .'i!Vd7 1 1
l2JbS �d8 1 gains the bishop-pair; e.g.,
12 e3 lZ:lxg3 13 hxg3 .i.g7 1 4 :b l lZ:la6
IS ..te2 lZ:lc7 = Jovanovich-Emma,
Thea 1 970, even though 12 e4 lZ:lxg3
13 fxg3 1 does give White dangerous
chances down the f-file, Bjel-Sorosi,
COIT. 1 978) 1 1 'ili'e4+ ( 1 1 lZ:lbS 'ili'b6)
45
1 l . ..'ti'e7 1 2 e3 ( 1 2 i.xd6?1 'iWxe4 1 3
lZ:lxe4 fS 1 4 i.xf8 fxe4 I S .i.xh6 :xh6
1 6 lZ:lxgS e3 1, Shadursky-Suetin, Vlad­
imir 1 962) 1 2 ...lZ:lxg3 1 3 hxg3 .i.g7 1 4
lZ:ld2 ( 1 4 'iii'c2 lZ:leS I S lZ:lxeS .i.xeS 1 6
.i.bS+ �f8 = , intending . . .a6 followed
by . . .bS, and answering 1 7 a4 with
17 ...�g7) 14 ...lZ:leS I S .i.bS+ �f8 1 6
lZ:lc4 :b8 ! ? 1 7 lZ:lxeS i.xeS 1 8 .i.d3 fS
1 9 'ili'a4 bS 20 'ili'c2 (20 .i.xbS a6)
20. . . f4 ! with an attack on both wings,
Gil Gonzales-Dolmatov, Barcelona
1 983 .
b) 1 0 lZ:ld2 can be a challenging
move in these lines. Here Black has
the reply 1O ...lZ:ld7 ! (anticipating lZ:lc4;
1O . . ..i.g7 transposes to Line A2) 1 1
lZ:lc4 lZ:lb6 1 2 e3 lZ:lxg3 1 3 hxg3 a6 (the
immediate 1 3 ... i.g7 is also possible)
14 a4 .i.g7 I S 'ti'd2 lZ:lxc4 16 .i.xc4
i.d7 17 as 'iWf6 ( 1 7 ...bS 1 8 axb6 'ili'xb6
= Agdestein; the two bishops and b­
file pressure compensate for Black's
weaknesses) 1 8 :a3 0-0-0 1 9 :b3,
Agdestein-Ljubojevic, Wijk aan Zee
1 988, and now Agdestein makes the
very interesting suggestion 1 9 ...�c7 !
with the idea . . . .i.c8, which absolutely
46 THE GAMBIT GUIDE TO THE MODERN BENONI
secures the king and prepares pawn­
breaks such as ... hS-h4 or .. .fS-f4.
10 ltJxg3
•••
This avoids 1O ... iLg7 1 1 iLbS+,
which has been played in a number of
top-flight games over the years, and is
generally acknowledged to be per­
fectly playable for Black (see ECO,
for example). But if you prefer to al­
low 1 1 iLbS+ (or transpose into this
position by 8 . . .iLg7 9 e3 gS 10 iLg3
ltJhS 1 1 iLbS+), the main line goes
1 1 . ..'it>f8 1 2 iLd3 ! ltJxg3 and then:
a) 1 3 fxg3 ! ? �e7 14 0-0 ltJd7 I S
iLfS ltJf6 1 6 ltJd2, and now the safe
1 6 . . . a6 1 7 �f3 'ito>g8 1 8 iLxc8 �xc8 1 9
11f2 �e8 20 �afl hS ! = led to a draw in
Yusupov-Gavrikov, USSR Ch (Frunze)
1 98 1 . Schneider suggests instead
1 6 ...�xe3+ ' ! ' 17 'ito>h l iLxfS 1 8 l:.xfS
�d3 "and White has no attack for the
sacrificed pawn", which is an exagger­
ation in view of 1 9 �f3 ! ; for example,
19 .. :iVd4 20 �c2 ltJxdS? ! 2 1 �fS or
1 9 .. :�g6 20 'iib3 b6 2 1 l:.afl .
b) 1 3 hxg3 ltJd7 14 �c2 �e7 I S
iLfS ( 1 S O-O?! h S 1 6 iLfS ltJeS +
Deze-Velimirovic, Sombor 1 972) and
now I S . . .ltJf6, I S ...ltJb6 and I S . . .11b8
all give Black a satisfactory game.
11 hxg3 iLg7 (D)
This position is characterized by
the trade-off between Black's weak­
nesses on the kingside and his bishop­
pair. Other factors inc1ude the bishop
on g7, which is particularly impres­
sive, and the potential insecurity of
Black's king. Now White has two
main moves:
All: 12 ltJd2 46
A12: 12 iLd3 47
w
1 2 iLbS+ is fairly pointless now;
e.g., 1 2 ...ltJd7 ( 1 2 . . .iL.d7 1 3 a4 iLxbS
14 axbS ltJd7 = ; 1 2 . . .'ito>f8 ! ? Psakhis)
1 3 �c2 a6 14 iL.d3 bS I S a4 c4 ! ? 1 6
iLfS b4 1 7 iLxd7+ iL.xd7 ! 1 8 ltJe4
�e7 1 9 �xc4 fS 20 ltJed2 iLxb2 2 1
.l:.bl iL.c3 + 22 ltJd4? �c8 (22 . . JlVe4 ! )
23 ltJc6 iLxc6 24 dxc6 'i¥e4 2 S 'iixe4+
fxe4 26 'iite2 dS + Plaskett-Ward, St
Helier 1 998.
Al l)
12 ltJd2 ltJd7! (D)
This anticipates ltJc4. 12 ...0-0 would
be Line A2, and 1 2 ... a6 1 3 a4 ltJd7 is
similar to 12 ... ltJd7; e.g., 1 4 ltJce4 ( 14
iLe2 ltJeS I S ltJc4 ltJxc4 16 iL.xc4
iLd7 = Osmanagic-Portisch, Sarajevo
1 962; 1 4 ltJc4 ltJeS I S ltJxeS iLxeS 1 6
iLd3 iLd7 1 7 0-0 0-0 1 8 'iVh5 iLg7 =
1 9 g4 ! ? �c8 20 iLe2 fS ! + Stigar­
Bjamason, Hamar 1 983) 14... ltJf6 I S
ltJxf6+ �xf6 1 6 ltJc4 0-0 1 7 .l:l.a2 ! ?
iLfS ( 1 7 . . .iL.d7 1 8 aS l1ac8 i s unc1ear ­
Cebalo) 1 8 as ltae8 1 9 iLe2 'iVe7 20
0-0 iLd7 = intending . . .fS and perhaps
... iL.bS at some point, Agzamov-Ceb­
al0, Vrsac 1 983.
 SYSTEMS WITH JLg5
W
13 'iVc2
Or:
a) 1 3 tiJc4 tiJeS ( 1 3 ... tiJb6) 14 tiJxe5
.ixe5 1 5 'iVc2 a6 1 6 a4 i.g7 1 7 i.d3
'V/IJe7 with equality, Bagirov-Savon,
Moscow tt 1 973.
b) 13 i.e2 'iVe7 1 4 g4 :b8 IS 0-0
�f8 1 6 a4 tiJf6 is 'unclear' - ECG.
Filip-E.Green, Nice OL 1 974 contin­
ued 17 :el ( 1 7 tiJc4 h5 1 8 gxhS g4 19
tLlbS ! ? ':'xhS ! , hitting d5) 1 7 ...hS 18
gxh5 g4 1 9 g3 'ilVe5 20 i.f1 'iVxhS 21
.ig2 and now Black could have tried
2 l . ..tiJd7 22 tiJb5 (22 tiJc4 tiJeS 23
tLlxeS i.xe5 24 'it>f1 a6) 22 ...tiJe5 23
tLlxd6 tiJd3 24 'iVc2 tiJxel 2S 1'hel b6
with the advantage.
13 tiJeS
•..
Or 1 3 ...'ilVe7 1 4 a4 tiJeS I S i.b5+
'it'f8 16 as h5 ! 1 7 a6 h4 1 8 gxh4 gxh4
19 axb7 i.xb7 20 l:ta4 i.c8 2 1 i.e2
Mb8 22 'itd 1 ! ? i.f6 23 'it>c l ?! :g8 ! 24
.if1 tiJg4 =+= Karpov-Gavrikov, Mazat­
Ian rpd 1 988.
The text-move ( 1 3 ...tiJeS) is Geller­
Malaniuk, USSR Ch (Moscow) 1 983,
which continued 1 4 i.b5+ ( 14 i.e2 =
Geller) 14 . . .i.d7 15 a4 0-0 ( ' = ' ECG)
47
1 6 i.e2 'ilVe7 1 7 l:Ia3 f5 ! with an equal
position.
A12)
12 i.d3
Probably the best move, concentrat­
ing on Black's weakened f5-square.
12 tiJd7 13 'iVc2 'iVe7 (D)
•.•
1 3 . . . a6 1 4 a4 .tIb8 is also played, but
I prefer the text-move.
W
14 a4
Or:
a) 14 i.f5 ! ? 0-0 I S 0-0 a6 16 a4
':'b8 1 7 as bS 1 8 axb6 :xb6 1 9 :a4
tiJe5 20 i.xc8 :!xc8 =+= Polaczek-de
Firmian, Philadelphia 1 989. The b-file
is more important than f5.
b) 14 tiJd2 tiJe5 1 5 i.fS i.xfS 1 6
'ilVxf5 'ilVd7 1 7 'i!Vxd7+ ( 1 7 tiJde4 'ite7 !
=+=; after the exchange of queens, ...fS
will come with tempo) 1 7 . . .Wxd7 1 8
'it>e2 fS 1 9 a4 l:the8 = Vilela-Ye Jiang­
chuan, Lucerne OL 1 982.
c) 14 O-O?! h5, when . . .h4 at some
point could prove very dangerous for
White.
14 a6 1S i.fS
.••
48 THE GAMBIT GUIDE TO THE MODERN BENONI
White wants to exchange the bishop
on c8 to secure the light squares, but
this costs time. Alternatives:
a) I S 0-0 0-0 (after I S . . . hS ' ! ' 1 6
.1fS g4, Boersma continues 1 7 ttJd2
h4, which looks good for Black, but 1 7
ttJh4 ! appears very strong) 1 6 :tab I
ttJeS 17 ttJxeS 'i!i'xeS 1 8 as l:tb8 = Hart­
ston-Nunn, British Ch playoff (Lon­
don) (3) 1 98 1 .
b ) I S ttJd2 ttJeS 1 6 .1e4 ! ? ( 1 6 .1fS
c4 ! ?) 1 6 .. J:tb8 ( 1 6... 0-0 =, intending
...fS) 1 7 as ..td7 1 8 0-0 hS ! 1 9 f3 ! h4
( 1 9 ...bS 20 axb6 l:.xb6 +) 20 g4 h3?!
2 1 g3 is unclear, Chetverik-Widera,
Fridek Mistek 1 995.
c) I S as ttJeS 1 6 ttJxeS �xeS 1 7
ttJa4 and now ECO gives 1 7 .. J:tb8 =.
Black could also continue 17 .....td7 1 8
lIbl .1xa4 ! 1 9 'i!i'xa4+ �e7 = .
15 ttJe5
•••
Black clarifies the situation imme­
diately. He reasons that fS can be con­
tested, and that his pawn-structure will
limit the mobility of White's knights.
This solution seems preferable to
I S . . .b6 ! ? 1 6 ttJd2 l:tb8 (ECO), but
I S ...ttJf6 16 as 0-0 17 ttJd2 ..td7 1 8 0-0
.1bS, as in Groszpeter-Bilek, Hungary
1 983, indicates another possible direc­
tion of play.
16 a5
This is considered the most dan­
gerous move. 16 ..txc8 l:txc8 17 �fS
�d7 (or 1 7 ...l:tb8) 1 8 'i!i'xd7+ ttJxd7
( 1 8 ... �xd7 =) 1 9 ttJd2 c4! 20 �e2 fS
2 1 as .1xc3 ! 22 bxc3 ttJf6 + Yuferov­
Kindermann, Naleczow 1 984.
16 .1xf5 17 �xf5 �d7
.•.
1 7 . . . ttJd7 may also equalize: 1 8
ttJd2 ( 1 8 l:ta4 ! ? �f6 ! ? with the idea
that 1 9 ttJe4 is met by 1 9 ...'i!i'xb2! 20
ttJxd6+ �e7 2 1 ttJxf7 .1c3+ 22 Wdl
�b3+ 23 �c2 'i!i'xdS+ with an unclear
position) was played in Groszpeter­
Kindermann, Budapest 1 985. Then
1 8 ...l:.b8 19 ttJc4 0-0 20 ttJa4 l:tfe8,
given as ';;\;' , should be OK for Black,
since ...ttJeS follows.
18 'i!i'c2 (D)
B
Equal, according to ECO.
18 0-0-0
•••
This is not the only move. For ex­
ample, Black could try 1 8 . . . l:tc8 1 9
ttJxeS ..txeS 20 ttJa4 l:tc7 2 1 e4 c4 22
l:ta3 ..td4 ! ? (22 ...�bS ! ? 23 ttJc3 'i!i'd7 !
= with the point 24 l:ta4 l:tcS) 23 0-0
�bS 24 ttJc3 'ilVcS ! 2S 'i!i'a4+ �d8,
when he has ideas like ...l:te8 or ...1:.f8
and .. .fS.
After the text-move ( 1 8 . . . 0-0-0),
Tukmakov-Agzamov, Erevan 1 982
continued 19 ttJxeS .1xeS 20 ttJa4 �b8
2 1 ttJb6, and now Black could have
tried 2 1 .. .'i!i'bS ! 22 ttJc4 (22 l:ta2 fS ! 23
ttJc4 { 23 'ii'a4? .1c3+ ! } 23 . . .hS +, in
view of 24 ttJxeS? dxeS 2S 'it'xfS??
'ii'b4+) 22 ... �b4+ (or 22 ... fS =) 23
 SYSTEMS WITH Jt..g5
'it'd l fS 24 l::tc l �g7 with no prob­
lems.
Playing to ensure the exchange of
White's bishop by . . . h6, . . . gS and
...ttJhS is Black's surest route to equal­
ity. Neither 1 2 lDd2 nor 1 2 i.d3
achieves any advantage in this line,
and Black can always keep the game
double-edged if he wants to.
A2)
8 �g7 (D)
...
I will cover this move in addition to
8 ... gS, into which it sometimes trans­
poses, both in order to give the reader
a good alternative and because so
many important games have been
played with it. As my presentation is
not exhaustive, the reader may wish to
consult other Benoni sources to inves­
tigate irregular moves and other de­
tails.
g e3
9 ttJd2 gS 1 0 i.g3 lDhS and now:
a) 1 1 e3 transposes to Line A I .
b) 1 1 'iVa4+ differs from 1 0 'iVa4+
in Line A l (note 'a' to White's 1 0th
49
move) since f8 is available. 1 1 ...'�f8
1 2 e3 and now 1 2 ...a6 1 3 'ili'c2 bS =
Sale-Zelcic, Solin 1 996, or 1 2 ... ttJxg3
1 3 hxg3 ttJd7 14 'iVc2 ttJeS = Uhl­
mann-Espig, Raach 1 969.
c) 1 1 ttJc4 ttJxg3 12 hxg3 0-0 1 3 e3
'iVe7 14 'iVc2 ( 1 4 i.eH td8 IS 0-0 ttJd7
16 a4 ttJeS = Botvinnik-Tal, Moscow
Wch (2) 1 960; 14 i.d3 is well an­
swered by 14 ... ttJa6 ! , continuing to
monitor fS, because I S ttJbS ! ? i.d7 !
1 6 ttJcxd6 c4! leaves White's knights
stranded, and intends 1 7 i.b 1 fS ! or 1 7
�xc4? i.xbS 1 8 ttJxbS 'iVb4+) 1 4... fS
I S i.e2 i.d7 1 6 a4 ttJa6 1 7 0-0 ttJb4
1 8 'iVd2 lh_lh Smejkal-Ftacnik, Prague
1 989. I'd slightly prefer Black because
of his two bishops.
9 0-0
••.
Now 9 ...gS 1 0 i.g3 ttJhS transposes
to the note to Black's 10th move in
Line A I . That is quite playable, but I
want to give the reader a safer alterna­
tive.
10 ttJd2 lDa6!? (D)
In response to White's rather slow
set-up (e3 and ttJd2), Black aims for
50 THE GAMBIT GUIDE TO THE MODERN BENONI
. . .tDc7 and . . .bS . This is a slightly un­
usual but respectable strategy.
11 i.c4
White has a number of reasonable
alternatives, probably of equivalent
value to the text-move:
a) 1 1 i.d3 tDc7 ! (threatening . . .gS
and . . .tDcxdS ; I think this is better than
the speculative 1 1 ...tDb4 ! ? 1 2 .i.c4 a6
1 3 a3 gS 1 4 i.g3 bS, O.Foisor-Suba,
Romania 1 983); e.g., 1 2 'iVb3 nb8 1 3
a4 ( 1 3 tDde4 bS 14 0-0 i.b7 ! =) 1 3 ...b6
1 4 0-0 a6 ! ? ( 1 4... i.b7) I S tDc4 bS 1 6
axbS axbS 1 7 tDaS i.d7 =.
b) 1 1 tDc4 tDc7 12 a4 b6 13 i.e2
i.a6 14 0-0 'it'd7 ! (coordinating Black's
pieces more efficiently than 1 4...i.xc4
I S i.xc4 a6 1 6 h3 ! 'ii'd 7, Ehlvest­
Bellon, Logroiio 1 99 1 ) I S h3 1:.ae8
1 6 i.g3 i.xc4 17 i.xc4 tDe4 = Law­
Nunn, London 1 977.
c) 1 1 i.e2 tDc7 12 0-0 bS ! ? 1 3
tDxbS ( 1 3 .i.xbS :b8 1 4 .i.d3 ! ? 1:bb2
I S tDc4 l:[b8 1 6 'iVc2 .i.a6 ! ? 1 7 nabl
1:.b4 with a type of dynamic equality;
both sides have many options, of
course) 1 3 . . .tDxbS 1 4 i.xbS nb8 I S
'ii'e2 g S 1 6 .i.g3 tDxdS = Bannik­
Suetin, Sochi 1 978. This game de­
serves further investigation, because
the same position could be reached in
variations with i.d3 and i.c4 as well.
See note 'a' to Black's 1 2th move.
11 tDc7 12 0-0
.•.
1 2 a4 nb8 1 3 0-0 seems well met
by 1 3 . . . b6, since White's slow devel­
opment gives Black time to play for
...a6, and he can also play . . .i.a6 in
some cases; e.g., 1 4 e4 ( 14 h3 a6 I S
'it'e2 ne8 with the ideas . . .i.b7 and
...'ii'd7) 14 ... a6 I S 'ii'f3? ! bS 1 6 axbS
axbS 1 7 i.d3? gS 1 8 i.g3 i.g4 1 9 'ii'e3
1:.e8.
12...ne8!?
The alternatives are interesting:
a) 1 2 . . .bS ! ? 1 3 tDxbS tDxbS 1 4
i.xbS transposes to note ' c ' to White's
1 1 th move.
b) 1 2. . . a6 1 3 a4 :b8 14 as bS I S
axb6 :xb6 compares well for Black
with the main line.
13 a4 .i.d7!?
Not bad, although again, 13 ...nb8
14 'iVe2 b6 ! is logical, since the c8-
bishop supports . . . a6, can pressure the
centre by ... i.b7, and may even have
occasion to exchange a knight on c4
after . . . i.a6. The main thing wrong
with White's position is that it's hard
to improve upon.
The text-move ( 1 3 ... i.d7) is Stem­
pin-Stoica, Polanica Zdroj 1 983, which
continued 14 'iVe2 nb8 I S :fc 1 ! ? a6
1 6 as bS 17 axb6 ':xb6 1 8 1:.a2 'ii'b 8
19 b3 'ii'b7? ! ( 1 9 ... tDbS ! = is prefera­
ble) 20 e4 tDbS 2 1 tDa4 tDd4 22 'ii'd3
i.xa4 ! 23 1ha4 tDbS, which is almost
equal, although Black is somewhat
tied down.
If one is going to play 8 . . .i.g7 in­
stead of 8 . . . gS , then the plan of
... tDa6-c7 is an unpretentious way to
achieve equality.
B)
6 e4 g6 (D)
7 tDf3
7 i.gS first is an interesting idea.
This move-order is neglected by every
leading source (except for a footnote
in ECO), since it tends to transpose if
play continues 7 . . .i.g7 8 tDf3 or 7 . . .h6
 SYSTEMS WITH ,i,g5
S .ih4 .ig7 9 tLJf3. But White can at­
tempt to play independently by, for
example, .id3 and/or 'iid 2:
a) 7 ... h6 S .ih4 i.. g7 is maybe the
less demanding option; e.g., 9 'iia4+ ! ?
(9 i..d3 0-0 1 0 tLJge2 l:leS 1 1 0-0 a6 1 2
a4 tLJbd7 1 3 l:lb 1 ! ? tLJeS 14 i..c2 i..d7
= Nardin-Savoia, Nova Gorica 1 999;
compare lines below) 9 ... i..d7 10 'iib3
WIIc7 1 1 i..c4 0-0 1 2 tLJge2 gS ( 1 2...a6
13 a4 gS 14 i.. g3 tLJhS, Uhlmann-Trin­
gov, Varna OL 1 962, has been given as
equal, but I S i..x d6! is too strong) 1 3
iog3 tLJhS = , in view of 1 4 i.. xd6
WIIxd6 I S 'iixb7 'iib6 16 'iixaS tLJa6.
b) After 7 ... i..g7, White has a num­
ber of independent options:
b l ) S i..d3 a6 (S ... O-O 9 'iid2 a6 1 0
a4 J:teS 1 1 tLJge2 tLJbd7 1 2 f4? h6 1 3
.ih4, Nikolaou-Petraki, Poros 1 995,
1 3 . . .tLJxe4 !) 9 a4 (9 tLJge2 ! ? bS 1 0 0-0
h6 1 1 i..h4 0-0 =) 9 . . .h6 10 i..h4 0-0 =
intending . . . tLJbd7-eS, and in some
cases, ... gS with ... tLJhS.
b2) S i..bS+ tLJbd7 (or S . . .i..d7) 9
f4 is very ambitious before White has
developed: 9 . . . 0-0 1 0 tLJf3 h6 1 1 i..h4
WIIaS ! ? 1 2 0-0, Celiz-Caramia, Acasusso
51
1 994, and the most direct course is
now 12 ... tLJxe4 ! 1 3 tLJxe4 'ilfxbS ; e.g.,
14 tLJxd6 'ilfxb2 I S l:tb l 'iixa2 16 i..e7
tLJf6 1 7 i.. xfS i..xfS I S tLJxcS l:lxc8 19
l1xb7 'ilfa6 20 l:lb l lIdS.
b3) 8 i..e2 h6 9 i.h4 (9 i..e3 a6 { or
9 ...0-0 followed by .. J:te8 and ...tLJa6}
10 a4 0-0 1 1 'iWd2 l:le8 12 f3 hS =; e.g.,
13 i.. gS 'ilfaS or 13 i..d l ? ! tLJbd7 14
tLJge2 tLJeS + - see ECO) 9 ... 0-0 and
now 10 'iid2 'iie7 is equal, while 1 0
tLJf3 transposes to Line C .
b4) 8 'ilfd2 0-0 9 i..e2 ( 9 i..h6 l:le8
10 i.. xg7 tLJxe4! ) 9 . . . a6 10 a4 'ii'aS in­
tending . . . bS and moves like . . .'iib4
and ... l:le8.
bS) 8 f4 'ifb6 ! ? 9 'ili'c2 0-0 10 tLJf3
l:le8 1 1 tLJd2 a6 ( l l . ..tLJxe4 ! 1 2 tLJcxe4
fS looks quite strong) 1 2 i..d3 tLJg4 1 3
tLJc4 'iic7 1 4 a4 ( 1 4 h3 tLJf6) 1 4. . .i..d 4!
= Goczo-Istrati, Kiskunhalas jr tt 1 995
(a game between two 1 2-year-old
girls !).
7 i..g7 8 i.. gS
•.•
Most of the theory cited here is
older, but it is still essential that Black
know what he's doing.
8 h6 (D)
•••
W
52 THE GAMBIT GUIDE TO THE MODERN BENONI
9 i.h4
9 i.f4 is tricky. 9 ... 0-0 (9 ... g5 10
�b5+! �f8 1 1 i.e3 ;!; Uhlmann-Golz,
Zinnowitz 1967) 1 0 tDd2 (a sort of
Knight's Tour; 10 i.e2 a6 1 1 a4 i.g4
1 2 tDd2 i.xe2 1 3 �xe2 tDh5 14 i.e3
tDd7 1 5 tDc4 tDb6 16 0-0 tDxc4 1 7
'iWxc4 �h4 ! was equal i n Tararykov­
Sergienko, Voronezh 1 998) 1 O...b6 ! ?
(better i s 1O...�e7 ! 1 1 i.e2 tDbd7 1 2
0-0 a6 1 3 a4 :b8; e.g., 1 4 tDc4 tDe8 1 5
a5 b5 1 6 axb6 tDxb6) 1 1 i.e2 i.a6 1 2
0-0 i.xe2 1 3 'ilt'xe2 a6 14 a4 tDh5 1 5
�e3 tDd7 1 6 �h l tDe5 ! ? 1 7 g4! ( 1 7 f4
'ifh4 ! 1 8 i.f2 �xf4 1 9 i.xc5 �h4 ! 20
�f2 �d8 =) 17 ...tDf6 1 8 h3 g5 1 9 f4
gxf4 20 i.xf4 tDfd7 2 1 tDf3 and with
the prospect of an eventual g5, White's
chances seem preferable, Kljako-Jur­
kovic, Zagreb . 1 998.
9...a6
9 ... g5 10 i.g3 tDh5 1 1 i.b5+ �f8
1 2 e5 ! is a famous attacking line,
which, even if defensible, causes too
much trouble for Black.
10 tDd2
Or:
a) 10 a4 g5 1 1 i.g3 tDh5 and now:
a l ) 12 i.e2 tDxg3 1 3 hxg3 tDd7 14
0-0 0-0 1 5 'ilt'c2 :b8 is very comfort­
able for Black, EMartinez-de Firmian,
Las Vegas 1 993. See also Line C,
where similar positions arise without
... a6 and a4.
a2) 1 2 tDd2 tDxg3 1 3 hxg3 tDd7 14
tDc4 tDe5 1 5 tDxe5 ( 1 5 tDe3 g4! 1 6
�e2 'iVg5 =+= ECO) 1 5 . . . i.xe5 with
equality - ECO.
b) 10 i.e2 g5 1 1 i.g3 tDh5 1 2 0-0
( 1 2 a4 is line ' a l ' ) 1 2 . . .b5 1 3 'iWc2
tDxg3 14 fxg3 tDd7 with two bishops
and no problems, Radzikowska-Moi­
seev, Katowice 1 99 1 . White should
only recapture with fxg3 when he gains
a concrete attack thereby.
10...b5 (D)
An interesting transposition arises
after 1 0...0-0 1 1 a4 tDbd7 1 2 i.e2 ( 1 2
tDc4 'iVe7 ! ? 1 3 i.d3 tDe5) 1 2. . .'iVe7 1 3
0-0, which reaches a known variation
from the Classical Main Line. Play
can continue 1 3 ...g5 ! ? 14 i.g3 tDe5 1 5
�c2 ! ? tDe8 1 6 f4 ! ? tDg6! 1 7 fxg5 hxg5
1 8 :ae l i.d4+ 1 9 �hl �g7, which is
unclear according to ECO. But White
may have better, and giving up f5 so
early is not to my taste.
w
11 �c2
White's two alternatives are impor­
tant, involved, and very instructive:
a) 1 1 a4 is often played: 1 l .. .b4 1 2
tDcbl 0-0 (or 1 2 ...g 5 1 3 i.g3 �e7 =,
an interesting game going 14 f3 tDh5
1 5 tDc4 tDxg3 1 6 hxg3 tDd7 17 tDbd2
i.e5 1 8 g4! i.g3+ 1 9 �e2 with ob­
scure prospects, Bertok-Matulovic,
Ljubljana 1 960; then 19 ... i.f4 20 a5
l:ta7 is one idea, intending ...tDe5 and
 SYSTEMS WITH i.g5
. . . .id7-b5 ; Black's alternative move
12 . . .'iie7 has also done well) 1 3 .id3
( 1 3 .ie2 l:te8 14 f3 g5 15 i.f2 tDh5 1 6
tDc4 a5 1 7 0-0 tDf4 + Lambert-Nunn,
London 1 977) 1 3 . . .l:te8 1 4 0-0 tDbd7
( l 4 . . .'iic7 is also effective; then the
move in ECO, 1 5 l:tel , is well met by
1 5 . . .tDg4! 1 6 tDc4 tDe5) 1 5 l:te l ( 1 5 f4
1iic7 1 6 'iif3, Sorin-Ginzburg, Buenos
Aires 1994, might be met by 16 . . .tDxd5
17 exd5 .ixb2 1 8 .l:.a2 l:te3 ! 1 9 'iixe3
.id4 20 'iixd4 cxd4; although White
has plenty of material, his weak d­
pawn and Black' s queenside advan­
tages result in a dynamic equality)
15 . . .tDe5 1 6 i.fl g5 1 7 i.g3 l:ta7 ! 1 8
l:'i.a2 l1ae7 1 9 b3, Alburt-Tukmakov,
Ashkhabad 1 978, and now the simple
19 ...tDg6! 20 f3 tDh5 2 1 i.f2 tDhf4
gives Black at least equality, with ideas
like . . .tDe5 and a well-timed . . .f5 in the
offing. The reader will notice this
. . .l:ta7-e7 manoeuvre throughout the
book.
b) 1 1 i.e2 tDbd7 ( 1 1 .. .0-0 1 2 0-0
tDbd7 transposes to line 'b2 ' ) and
now:
b l ) 1 2 'iic2 c4 (or 1 2 . . . l:tb8, to an­
swer 1 3 a4 with 1 3 ...c4 - Nunn) 1 3
a4 ! ? l:tb8 1 4 axb5 axb5 1 5 b4 ! ? cxb3
16 'iixb3 0-0 17 0-0 tDc5 1 8 1Wb4?
tDxd5 ! 1 9 .ixd8 tDxb4 20 .ic7 l:tb7 2 1
i.xd6 i.xc3 22 i.xc5 l:td8 + Zsu.Pol­
gar-Suba, Dortmund 1 985.
b2) 1 2 0-0 0-0 1 3 a4 ( 1 3 'iic2 c4 14
f4 l:te8 15 l:tael tDc5 16 �h l i.d7 =
Sigurjonsson-Y.Grtinfeld, Lone Pine
1979) 1 3 . . .b4 14 tDcbl l:te8 1 5 f4 ! ?
WIIc7 1 6 .if3 c4 ! ( a bold counterat­
tack) 1 7 e5 c3 1 8 bxc3 ( 1 8 exf6 c2 1 9
WIIc l tDxf6 20 tDb3 cxbl'ii 2 1 l:txbl
53
'ii'xc l 22 l:tbxc 1 .id7 +) 18 . . .bxc3 1 9
'iic l ! ? tDxd5 ! 20 .ixd5 1Wc5+ 2 1 �h l
'iixd5 22 tDxc3 1We6 + A.Mikhalev­
ski-Psakhis, Israeli Ch (Jerusalem)
1 996.
11...0-0
The normal move, but not necessar­
ily the easiest one:
a) l 1 . . .g5 1 2 .ig3 tDh5 1 3 a4 b4 1 4
tDd l tDxg3 1 5 hxg3 tDd7 1 6 tDe3 tDe5
with no special problems ( 1 7 f4 tDg6 !),
Golz-Soos, Polanica Zdroj 1 968.
b) l l .. .tDbd7 12 a4 b4 13 tDd l 0-0
14 .ie2 (this position can also arise
via 7 tDf3 .ig7 8 .ie2 0-0 9 .ig5, i.e.
Line C) 14 . . .l:te8 15 tDe3 b3 ! 16 'iWbl
l:tb8 17 a5 ! ( 1 7 0-0 l:.b4; 1 7 tDec4
tDe5 1 8 tDxe5 l1xe5 =) 17 . . .l:tb4! 1 8
tDec4 tDe5 1 9 tDb6 g5 20 .ig3 tDxe4!
2 1 tDxe4 .if5 22 f3 tDxf3+ ! 23 gxf3,
Calderin-Gi.Hernandez, Merida Torre
mem 1 997. Now Black should proba­
bly settle for 23 . . . .ixe4 24 fxe4 l:tbxe4
25 0-0 l:txe2 26 'iVf5 ! �8e7 with a
small advantage.
12 a4 b4 13 tDdl .l:.e8 14 tDe3 (D)
14 .ie2 is another position that can
arise from Line C (7 tDf3 .ig7 8 i.e2
0-0 9 .ig5). Then among other moves
(such as 14 . . .tDbd7, which transposes
to note 'b' to Black's 1 1 th move), a fun
line is 1 4 . . .l::t a7 ! ?, with the usual idea
of .. .l::t ae7, but also preparing to meet
1 5 tDe3 with 1 5 . . .b3 ! 1 6 �xb3 ( 1 6
�bl g5 1 7 .ig3 l1ae7 1 8 f3 tDh5 1 9
.if2 f5 ! =) 1 6 . . .l1b7 1 7 'iWc2 g5 1 8
i.g3 tDxe4 ! 1 9 tDxe4 .l:.xb2 20 'iid3
l:txe4 ! 2 1 'iixe4 .ic3+ 22 �fl l:tb4 23
tDc4 .ixal 24 .id3 f5 25 'iie l .if6,
when Black has a clear advantage,
S .Mohr-Perenyi, Budapest 1 988.
54 THE GAMBIT GUIDE TO THE MODERN BENONI
14...b3!?
This is a typical diversionary ma­
noeuvre. 14 . . .tZJbd7 seems safer; e.g.,
IS i.d3 ( 1 S i.e2 b3) l S ... tZJeS ! ? (or
l S . . :tWc7 ! ? 1 6 0-0 tZJeS 1 7 tZJec4 t
Chen De-Wang Zili, Chinese Ch 1 987)
16 i.e2 ':'b8 ! ? = with the idea . . . b3;
e.g., 17 0-0 ( 1 7 b3 gS 1 8 i.g3 tZJg6 1 9
0-0 tZJf4 =) 1 7 ...b 3 1 8 tZJxb3 g S 1 9
i.g3 tZJxe4.
15 �d3
I S �b1 gS (perhaps l S . . .tZJbd7 ! ?
1 6 i.c4 { 1 6 f3 ':'b8 ! ? } 1 6 ...'i!VaS 1 7 f3
.l:i.b8 1 8 ':'a3 tZJeS ! 1 9 1hb3 i.d7 could
be tried, with pressure) 16 i.g3 tZJhS
1 7 i.d3 (Suba gives 1 7 i.c4 tZJf4 1 8
0-0 "tWf6, although White may well
have an edge) 17 . . . tZJf4 1 8 0-0 tZJd7 ! ?
1 9 tZJfS tZJeS 20 tZJxg7 �xg7 2 1 i.c4
"tWaS ! ? and from here Spiridonov-Suba,
Bajmok 1 980 was eventually drawn,
but this whole game is hard to trust.
IS... tZJbd7 16 f3
1 6 ':'a3 is answered by 1 6 ...:tb8 1 7
':'xb3 ':'xb3 1 8 "it'xb3 gS .
16...l:.b8 17 tZJec4 tZJb6
F.Portisch-D.Cramling, Reg­
liz-liz
gio Emilia 1 979. I think that this is
indeed about equal, since White again
has trouble rounding up the b-pawn
without tactical repercussions; for ex­
ample:
a) 1 8 �xb3 �c7 1 9 tZJxb6 ( 1 9 tZJe3
i.d7 ! 20 "it'c2 tZJbxdS !) 1 9 . . ..l:i.xb6 20
�a2 tZJxe4! 2 1 fxe4 ':'xb2 and .. :�aS .
b) 1 8 ':'a3 tZJxc4 1 9 tZJxc4 .i:tb4 ! 20
as (20 i.g3 tZJhS; 20 i.e2 as 2 1 0-0
i.a6 22 l:tc 1 �d7 =) 20. . .i.fS ! 21 i.e2
(2 1 g4 i.xe4 ! 22 fxe4 gS) 2 l . . .gS 22
exfS (22 i.g3 tZJxe4) 22 ...gxh4 with a
nice attack in view of 23 0-0 tZJxdS ! .
The . . .b3 idea is an essential Benoni
resource, although in this particular
case White's option of IS �bl worries
me (at least in theory). For those who
want something else, Black's 1 1 th­
move options are sound alternatives,
and 14 ...tZJbd7 can also be investigated.
C)
6 e4 g6 7 tZJf3 i.g7 8 i.e2 0-0 9
i. gS
This is the old main line. It is in
good theoretical shape from Black's
point of view, but unfortunately a lot
of analysis is required to prove that.
9...h6 (D)
10 i.h4
Or:
a) 1 0 i.f4 has been played fairly
often, but as opposed to the last sec­
tion, . . .bS ideas tend to be good, be­
cause i.xbS will occur at the loss of a
tempo. This yields many options for
Black containing instructive Benoni
themes:
aI) 1 O . . . a6 1 1 a4 i.g4 transposes
to a type of Modem/Classical Main
Line - see Chapter 9.
 SYSTEMS WITH iLg5
a2) 1 O... gS 1 1 �e3 bS ! ( l l . ..lbg4
1 2 �d2 fS =) 1 2 �xbS lbxe4 1 3 lbxe4
�aS+ 14 'iWd2 'iWxbS IS lbxd6 'iWa6 !
( l S ...'iWxb2) and again I like Black's
compensation, with the possible con­
tinuation 1 6 �xcS lbd7 1 7 �a3 lbeS
1 8 :td 1 lbxf3+ 1 9 gxf3 l1d8 20 h4
I!xd6 2 1 �xd6 'iWxd6 22 hxgS �a6,
which is at least equal.
a3) A restrained approach for Black
is 1 O . . .lba6 1 1 lbd2 lbc7 1 2 lbc4
lbfe8 1 3 a4 .l:.b8 14 0-0 b6 I S lbbS
lbxbS ! 1 6 axbS :tb7 = Barus-Khous­
seinov, Kuala Lumpur 1 993.
a4) 1O ... bS 1 1 �xbS ( 1 1 �xd6 =
was given by Tal; then 1 1 ...'iWxd6 1 2
eS 'iWb6 1 3 exf6 �xf6 i s rather prom­
ising for Black, according to Schnei­
der's analysis; 1 1 lbd2 a6 1 2 0-0 :te8
= Gligoric-Malich, Munich OL 1 9S8)
1 1 ... lbxe4 1 2 lbxe4 'iWaS+ leads to
equality:
a4 1 ) 1 3 lbc3 �xc3+ 14 bxc3 'iWxbS
IS �xh6 :te8+ 1 6 �e3 .i.a6 1 7 'iWd2
lbd7 =.
a42) 1 3 lbfd2 'iWxbS 14 lbxd6 'iWxb2
I S :tc l �a6 ! ? = (or I s ...lbd7 =) 1 6
�c2 "ii'f6 =.
55
a43) 13 'iWd2 'iWxbS 14 lbxd6 "ii'a 6!
IS :tc 1 gS 16 .i.g3 lbd7 =; e.g., 1 7 h4
( 1 7 'iWe2 'iWaS+) 1 7 ... g4 18 lbh2 lbeS
1 9 �xeS �xeS 20 lbc4 �d4 ! intend­
ing ...:te8+.
b) 10 �e3 could have the same an­
swer: 1 O... bS ! 1 1 �xbS lbxe4 1 2
lbxe4 'iWaS+ 1 3 'iWd2 ( 1 3 lbc3 .i.xc3+
14 bxc3 'iWxbS I S 'iWb3 �a6! Psakhis)
1 3 ...'iWxbS 14 lbxd6 'iWxb2 I S :te l is
given as unclear in ECO. There might
follow something like IS ...'iWxd2+ 1 6
lbxd2 .i.a6 1 7 lhcS ( 1 7 �xcS lbd7)
1 7 ... fS ! 1 8 lb2c4 :td8 1 9 :tc7 �f8 ! 20
lbf7 :td7 2 1 ':xd7 lbxd7, with at least
equality.
10 g5
..•
1 0 ... a6 1 1 lbd2 bS 1 2 0-0 lbbd7
transposes to note 'b2' to White's 1 1 th
move in Line B .
11 �g3 lbh5 (D)
1 1 ...bS is also played and consid­
ered equal or unclear; see ECO, for ex­
ample.
12 lbd2
This move is more pointed than 1 2
0-0 and now:
56 THE GAMBIT GUIDE TO THE MODERN BENONI
a) 1 2 ... ltJxg3 1 3 hxg3 f5 14 exf5
.Jtxf5 1 5 ltJd2 ltJd7 is equal:
a 1 ) 16 ltJc4 ltJe5 17 ltJxe5 i.xe5 1 8
i.d3 'ti'f6 1 9 i.xf5 'iWxf5 20 'iWd2,
Pfleger-Hindle, Hastings 1 964/5, and
now Black can keep a clear edge by
20...':'ae8 2 1 ':ael a6 22 a4 i.d4.
a2) A recent example proceeded
1 6 i.g4 i.g6! 1 7 i.e6+ �h8 1 8 ltJde4
ltJf6 1 9 'iWe2 b5 ! ? (or 1 9 ...ltJxe4 ! 20
ltJxe4 .Jtd4 =1=) 20 ltJd2 a6 and the two
bishops and queenside majority give
Black the advantage, Hartung Niel­
sen-H.Olsen, Herlev 1 997.
b) 1 2 ...f5 is similar, with a unique
situation arising after 1 3 ltJd2 ltJxg3 14
fxg3 ! ? 'iWe7 1 5 exf5 i.xf5 1 6 i.g4 ! ?
i.xg4 1 7 ':xf8+ 'iitxf8 ! 1 8 'iWxg4 ltJd7
1 9 ':f1 + 'iitg 8 =1= 20 'iie6+ 'iWxe6 2 1
dxe6 i.d4+ 22 c;t>h 1 i.xc3 23 bxc3
ltJf8 24 ':'el ':'e8 =1= Aaron-Robatsch,
Varna OL 1 962.
12 ltJxg3 13 hxg3 fS
•••
This active solution depends upon
the strength of the two bishops and the
f-file to offset the light-square weak­
nesses created. I think that this is an
exciting way of proceeding.
The main line in most sources, and
the choice of most players, is 1 3 ... ltJd7
( 1 3 ... a6 1 4 a4 ltJd7 is also played),
which probably suffices for equality,
but requires a more delicate hand with
the move-order. One opinion of mine
that the reader may find useful if he
wishes to investigate the 1 3 ... ltJd7
line, is that after 1 4 ltJc4 'iWe7 1 5 'iWc2
(the main line), I seriously doubt that
the accepted move 1 5 ... ltJf6 com­
pletely equalizes, and recommend in­
stead 1 5 ...ltJe5 1 6 ltJe3 .Jtd7 ! ? 1 7 a4,
which is supposed to favour White, but
here I think that 1 7 .. J:tfb8 ! ? should be
considered: 1 8 f4 ( 1 8 a5 b5 1 9 axb6
axb6 20 0-0 ltJg6 =; 1 8 i.b5 i.c8; 1 8
ltJb5 a6 1 9 ltJc7 ':a7 =1=) 1 8. . .ltJg6,
which ultimately should be equal.
14 exfS
Otherwise:
a) White should avoid 14 O-O?! fxe4
( 1 4. . .f4 ! ? 1 5 gxf4 ':'xf4 = ) 1 5 ltJdxe4
i.f5 1 6 i.g4 ! ? i.g6! 1 7 i.e6+ 'it>h8 =1=
intending . . . i.d4 and/or . . .ltJa6-c7.
b) Black can answer 14 ltJc4 with
14 ...'iWe7 or 14 . . .ltJa6 1 5 exf5 .Jtxf5 1 6
0-0 'iWf6 ! - Psakhis.
14 i.xfS 1S ltJc4 (D)
•••
The most ambitious try. 1 5 i.g4?!
makes no sense in view of 15 ...'iWe7+
16 'it>f1 'iWd7 =1=, and after 1 5 0-0 ltJd7
1 6 ltJc4 ltJe5 White should prefer 1 7
ltJe3 i.g6 = , rather than 1 7 ltJxe5 ? !
i.xe5 1 8 i.d3 'iWf6 =1= Pfleger-Hindle,
Hastings 1 964/5 .
B
IS 'ti'e7
.••
Especially if you're uncomfortable
with ltJb5 ideas, 1 5 ... a6 ( ' ! ' Schneider)
is an important alternative:
 SYSTEMS WITH .i.,g5
a) After 1 6 a4 Wie7 there are two
continuations, one resembling the main
line and the other (line 'a2') taking a
unique course:
a l ) 17 ttJe3 ttJd7 1 8 ttJxf5 lhf5 1 9
0-0 ttJe5 ( 1 9 ... .i.d4 ! ? 20 .i.g4) 20 SLg4
ttJxg4 21 Wixg4 '=' Wade; this is too
modest an assessment, as 2 1 .. .':'af8 !
22 l:tael Wif7 is extremely strong; e.g.,
23 f3 .i.d4+ 24 'iti>h l h5 25 Wie4 .i.e5,
etc.
a2) 1 7 ttJb6 l:.a7 1 8 0-0 ( . . .ttJd7
follows in any case) 1 8 ...ttJd7 1 9 ttJxd7
( 1 9 a5 ttJxb6 20 axb6 l:.aa8 2 1 .i.g4
�d4!) 1 9 .. :iWxd7 20 a5 l:.aa8 2 1 ttJa4?!
(probably not best, but ... .i.d4 fol­
lowed by doubling on the f-file was a
serious problem) 2 1 . ..l:.ae8 22 ttJb6
Wic7 23 ttJc4? (23 'iIId2 .i.d4 24 SLh5
+) 23 ... .i.d4 24 .i.d3 .i.xd3 25 Wixd3
lif6 26 ttJe3 Wih7 ! + Bilek-L.Evans,
Amsterdam 1 964.
This interpolation of 1 5 ... a6 16 a4
is potentially a significant improve­
ment for Black. White has to look for a
more forcing idea, namely, the imme­
diate ...
b) 1 6 ttJe3. This move is surpris­
ingly unmentioned. Black has a vari­
ety of moves that are hard to assess,
especially when opposite-coloured
bishops arise. Here is a small subset of
possible continuations. 1 6 . . .Wif6 (D)
( l 6 . . . ttJd7 1 7 ttJxf5 l:.xf5 1 8 Wic2 ! ?
J::te5 19 0-0-0 b 5 ! ?) with these possi­
bilities:
b l ) 1 7 .i.g4 .i.h7 ! 18 .i.e6+ 'iti>h8
19 ttJg4 'iIIg6 20 Wie2 (20 a4 ttJc6 ! ; 20
0-0 b5 2 1 l:.el .l:ta7) 20... ttJc6! 2 1 dxc6
l:.ae8 22 cxb7 ':'xe6 23 ttJe3 Wif7 ! and
the bishops will be very strong.
57
b2) 17 ttJxf5 'iIIxf5 1 8 0-0 .i.d4 (or
1 8 ... b5 19 .i.g4 'iIIg6 20 .i.e6+ 'iti>h8
{ intending . . .l:.a7 and ...ttJe5 } 21 f4 ! ?
b4 2 2 f5 'iWf6 with good play) 1 9 .i.g4
'iIIe5 and now:
b2 1 ) 20 Wid2 b5 2 1 l:tael Wig7 22
':'e6 ttJd7 23 l:txd6 ttJe5 24 .i.e6+ 'iti>h8
with interesting play in view of ...ttJc4
or . . . c4 with . . . .i.c5 and/or . . .ttJd3.
Against the logical 25 Wic2 ! ?, Black
has 25 . . .l:ta7 ! ?, when 26 .i.f5 ? fails to
26 . . ..:.af7 27 g4 ttJxg4 ! . 26 ttJe4 is
better, when 26... .i.xb2 ! , while messy,
seems to be at least equal for Black;
e.g., 27 ttJxc5 .i.a3 ! ? or 27 Wixc5 ttJc4.
b22) 20 .i.e6+ 'iti>h8 21 Wid2 b5 22
l:.ael Wig7 23 l:.e2 ttJd7 24 ttJe4 'iIIe7 =
intending . . . ttJe5, when the power of
the d4-bishop matches that on e6.
16 ttJe3
16 0-0 prepares .i.g4 and gives up
pretensions down the h-file: 1 6. . .ttJd7
( l 6... ttJa6 1 7 .i.g4 ttJc7 might be a safe
alternative) 1 7 ttJb5 ! ? (one of those
cases I referred to in the previous note;
the 'normal' move is 17 .i.g4 SLxg4 1 8
Wixg4 and instead of Wade's 1 8 . . .ttJe5
'=', when the reply 1 9 .l:tael ! is unclear,
58 THE GAMBIT GUIDE TO THE MODERN BENONI
1 8 . . .1:tae8 ! + keeps ideas like . .. a6 and
....i.d4 alive) 1 7 ... .i.e4 ! (a Fritz 6 dis­
covery) 1 8 .i.g4 ( 1 8 f3? a6 1 9 lLibxd6
.i.d4+ 20 �h l .i.xd5 +; 1 8 lLibxd6
.i.xb2 1 9 lLixb2 'iWxd6 +; 1 8 lLicxd6
.i.xb2 1 9 lLixe4 .i.xa 1 20 'iic 2! .i.d4 ! ?,
and White has some compensation,
but at the very least, Black can give
back the exchange; e.g., 2 1 lLic7 lLif6!?
22 lLixf6+ 'iixf6 23 lLixa8 lha8 =)
1 8 . . . lLie5 1 9 .i.e6+ �h8 20 lLibxd6
lLixc4 2 1 lLixc4 b5 22 'iVe l ! (Fritz
again; 22 l:te l .i.h7 23 lLie5 l:txf2! 24
�xf2 .i.xe5 25 'iVh5 .i.d4+ is difficult
for White, as { perhaps? ! } is 22 lLid2
.i.d3 23 l:te l .i.xb2 ! ? 24 .i.f5 'iVxe l +
25 'fixe l lbf5) 22. . .bxc4 23 'iWxe4
.i.xb2 24 l:tab l .i.d4 =.
Complications like these motivate
the 1 5 ... a6 suggestion above.
16 lLid7 17 g4?!
.•.
Ultimately more of a liability than a
benefit. Alternatives:
a) 1 7 .i.d3 ? ! .i.d4! ( l 7 ... i.xd3 1 8
'ili'xd3 lLie5 i s also satisfactory, in view
of 1 9 'iic2? lLig4! 20 0-0 l:txf2! -+ or
1 9 'iVe2 c4 20 0-0 ':'ac8) 1 8 .i.xf5 ( 1 8
g4 .i.xd3 1 9 'iVxd3 lLie5 20 'iie2 l:hf2 !)
18 . . Jbf5 1 9 'iVe2 (19 'iYh5 l:tf6 +; 1 9
"it'd3 ':'af8 20 lLidl ? lLie5 with a win­
ning game for Black) 1 9 .. J:te8 20
lLicd l (20 g4 ':'e5 2 1 lLicd l lLif6 ! 22
':'xh6 lLixd5 -+) 20 .. :iVf8 ! 2 1 0-0 l:tfe5
+.
b) 17 lLixf5 ! ':'xf5 1 8 0-0 lLie5 =
Wade. Play might continue 1 9 �d2
':'af8 20 ':'ael 'iVd8 =, when White
mustn't overextend by 2 1 f4? due to
2 l .. .gxf4 22 gxf4 'iVh4 ! with an attack
and a clear advantage.
17 .i.g6 18 .i.d3 .i.xd3 19 'iWxd3
.•.
lLie5 20 'iWe2 ':'f4!? 21 0-0-0 'iWf8
A clearer course is 2 1 . ..lLixg4 22
lLixg4 'iVxe2 23 lLixh6+ i.xh6 24
lLixe2 ':'f6 25 lLig3 ':'e8 + .
22 f3 ':'e8 23 lLie4 c4 24 �bl
Avoiding 24 lLif5 lLixg4! (24...lLid3+
25 ':'xd3 ! is unclear) 25 lLixg7 Wixg7
+.
The text-move (24 �bl ) is EPor­
tisch-Bilek, Zalaegerszeg 1 969. Black
has an obvious advantage. The game
continued 24...c3?! (better is 24...lLid3 !
with the point 25 lLixc4 lLixb2 ! 26
lLixb2 ':'exe4 !) 25 bxc3 b5 ! 26 ':'he l
(26 ':'hfl would prevent the following
idea, but it's not necessary) 26 ...b4 27
c4? (27 cxb4 ! lLixf3 ! 28 gxf3 ':'exe4
29 fxe4 "it'f6 30 ':'c l 1:txe4! is unclear,
but probably a perpetual after 3 1 ':'c8+
�h7 32 ':'c4) 27 . . .lLixf3 ! 28 gxf3
':'exe4 29 fxe4 "it'f6 ! (29 ... ':'f2 30 lLif5 !
':'xe2 3 1 ':'xe2) 30 ':'d3? (but 30 'iVd2
':'f2 3 1 "it'd4 'iVxd4 32 ':'xd4 .i.xd4
won't last long: 33 lLif5 { 3 3 lLic2 .i.c3
and even ... a5-a4 and ...b3 is possible }
33 ... .i.e5 34 ':'h l ':'b2+ 35 �c l 1ha2
with ...b3 to come) 30...':'f2 3 1 'iVxf2
'iVaI + 0- 1 . A fine game, and typical of
the dynamism of this variation.
The systems of development with
.i.g5 cannot be taken lightly. As al­
ways, Black must continue aggres­
sively if he is to counteract White's
attempt for a bind. Fortunately, the ac­
quisition of the bishop-pair and the en­
hanced power of the g7-bishop (once
his dark-square counterpart is gone)
ensure Black lively and satisfactory
play.
4 The Kn ight's Tou r Va riation
1 d4 tDf6 2 c4 c5 3 d5 e6 4 tDc3 exd5 5
cxd5 d6 6 tDf3 g6 7 tDd2 (D)
B games to describe the theoretical ter­
7 tDd2 introduces the Knight's Tour
Variation, so named in view of the
early excursion of the f3-knight via d2
to c4. As discussed in Chapter 1 0,
White can also use this move-order to
get to the Classical Variation of the
Benoni by 7 tDd2 iLg7 8 e4 0-0 9 0-0.
In the variation before us, White's ba­
sic idea is to exert direct influence on
j6 by, for example, tDc4, iLf4, and
sometimes tDb5. As Black will soon
commit his bishop to g7, the d6-pawn
can be somewhat difficult to defend.
In 1 985, I wrote a repertoire book
that in part advocated the use of 7
ctJd2 iLg7 8 ttJc4 0-0 9 iLg5. Within a
few years, Black had solved most of
his problems in that line, and I was
surprised to find that the theory of this
variation has changed very little since.
The Knight' s Tour has become rela­
tively rare at top level and Benoni
sources tend to use the same older
rain. Nevertheless, the traditional main
lines are under-analysed for both sides,
so there is room for creativity.
After 7 tDd2, I will look at:
A: 7 tDbd7
••• 59
B: 7 iLg7
••• 64
A)
7 tDbd7 (D)
•••
w
A straightforward and solid defence.
Black makes a natural move and antic­
ipates responding to White's tDc4 by
...tDb6. Although it is not mentioned
by NCO or MCO-14, I am struck by
60 THE GAMBIT GUIDE TO THE MODERN BENONI
how effective this move is, and how
Black has two routes to equality at
several critical junctures. The reader
will especially like this solution if he
wants to bypass the complications af­
ter 7 . . . J..g7 8 liJc4 0-0, dealt with in
Line B.
8 e4
To the extent that theory concerns
itself with 7 . . .liJbd7, this is considered
the most challenging move. It intends,
after liJc4, to meet . . .liJb6 with liJe3
and afterwards, with a4-aS . The natu­
ral alternative 8 liJc4 liJb6 is a frequent
choice in practice, but presents Black
with no serious difficulties ifhe under­
stands the ideas:
a) 9 liJe3 is logical but a little awk­
ward. White hopes for a4-aS and a re­
turn to c4, but this allows Black to
reorganize: 9 . . .J.. g7 10 a4 J..d7 1 1 as
liJc8 1 2 liJc4 'iile7 1 3 J..gS h6 1 4 J..h4
0-0 ( 1 4. . . gS I S J..g3 liJhS =) I S e3 �b8
('=' Psakhis and ECO), Whiteley­
Schweda, Cappelle la Grande 1 992.
Play might continue 16 J..e2 bS 1 7
axb6 axb6 1 8 0-0 bS 1 9 liJaS?! ( 1 9
'iilb3 liJb6 20 liJxb6 .l:!.xb6 2 1 .l:!.a7 .l:!.fb8
22 �fal b4 23 liJbl gS 24 J.. g3 liJe4 2S
J..g4 1:I6b7 =) 1 9 ...b4 20 liJbl 'iVe4 2 1
J..xf6 J..xf6 with a slight advantage for
Black.
b) 9 liJxb6 'iVxb6 1 0 e4 J.. g7 1 1
J..bS+ liJd7 ! ? ( l l .. .J..d7 =) 1 2 'it'a4 a6
1 3 J..d2 "WIc7 1 4 0-0 0-0 I S .te2 ':'b8 =
Al Kazzaz-Chekhov, Caracas U-26
Wcht 1 976.
c) 9 e4 liJxc4 10 J..xc4 J..g7 1 1 0-0
0-0 (D) and now:
c 1 ) 1 2 J..f4 with some informative
examples:
w
c l l ) 1 2 . . . �e8 1 3 J..d3 a6 14 a4
liJhS ! ( 1 4. . . c4! ? I S J..c2 1:Ib8 1 6 as bS
1 7 axb6 'iilxb6 1 8 J..e3 'ii'd8 ;!; Foguel­
man-Seidler, Buenos Aires 1 976) I S
J..e3 ( I S J..d2? ! 'iVh4 ! ) I S . . .J..eS 1 6
�d2 ( 1 6 f4 J..xc3 1 7 bxc3 fS i s a little
better for Black) 1 6. . . J..d7 1 7 as ( 1 7
f4? ! J.. xc3 1 8 'it'xc3 ! liJf6 1 9 'iilc2 bS
+) 1 7 . . .'iilh4 ! 1 8 f4 J..g7 1 9 �el 'iilxel
20 l:taxel fS intending to meet 2 1
exfS?! (2 1 J..d2 J..d4+ 22 'it>h l liJf6 ! )
with 2 l . ..1:Ixe3 ! 22 lhe3 J..d4 2 3 liJd l
.l:!.e8 :t.
e l 2) 12 ... a6 1 3 a4 liJhS ( 1 3 . . .liJg4
Nunn) 14 J..e3 �e8 (Kapengut sug­
gests 14 . . . fS ! ? I S exfS .txfS with the
point that 1 6 g4?! is met by 1 6. . .'iilh4)
I S 'ilVd2 1:Ib8 1 6 �fel 'it'd7 ! 17 as bS
1 8 axb6 ':xb6 19 l:tae l 'it'b7 with an
equal position, Gligoric-Trifunovic,
Sombor 1 9S7.
c2) 12 J..gS h6 ( l 2 ... a6! is easier:
1 3 a4 J..d7 ! ? { or 1 3 .. J:te8 14 'iild2 h6!
I S J..xh6 liJxe4 = } 14 'iild2 bS ! I S axbS
axbS 1 6 J..xbS J.. xbS 1 7 liJxbS Ihal
1 8 lhal liJxe4 19 J.. xd8 liJxd2 is
equal, F.Portisch-Hodgson, Wijk aan
Zee 1 985) 1 3 .th4 gS ! ? (or 1 3 . . . a6 1 4
 THE KNIGHT'S TOUR VARIATION
a4 .i.d7, with the same idea as 1 2 ...a6)
14 .i.g3 a6 I S a4 Ite8 and now:
c2 1 ) 1 6 �el 4Jg4 17 .i.e2 4JeS 1 8
�d2 4Jg6 1 9 i.hS .l:lb8 20 .i.xg6 fxg6
21 f4, Barbero-A.Schneider, De1men­
horst 1 986, and now Hebert suggests
21 ...gxf4 ! 22 �xf4 �eS 23 l:tfl 'iVd7
24 �f2 bS 2S l:tafl b4 26 4Jd l .i.b7 27
ctJe3 �be8 as unclear.
c22) Kapengut analyses 1 6 .i.d3
ctJg4 1 7 h3 4JeS 1 8 f4 4Jxd3 1 9 �xd3
�b6 20 �h2 'iVxb2 ! ? 2 1 �ac 1 and
now gives 2 1 . . .c4! 22 �xc4 i.xh3
with "a sharp position with equal
chances"; Black holds on in the end­
ing after 23 �xh3 11ac8 24 'iVa2, etc.
But 2 l . ..g4 ! 22 hxg4 i.. xg4 appears
better, having in mind 23 fS? i.xc3 ! +.
8 ..tg7 9 4Jc4 (D)
.•.
For 9 i.e2 0-0 10 0-0 �e8, see Line
A of Chapter 10.
9 .i.d3 0-0 10 0-0 is an odd move­
order which may transpose to the main
line via 1 O ...�e8 1 1 4Jc4 4Jb6 1 2 4Je3,
but can also have independent value if,
for example, Black deviates now:
1O. . . 4JeS (or 1 0... a6 1 1 a4 �e8 1 2 f4
Mb8 1 3 h3 'iVc7 1 4 11f3 c4 I S 4Jxc4 bS
16 axbS axbS 1 7 4Jxd6 ! ? �xd6 1 8 eS
l:.xeS 1 9 fxeS 4JxeS with surprising
but speculative compensation for the
exchange, Sochor-Simacek, Moravka
jr 1 994) 1 1 i.e2 gS ! ? (compare the
Classical Main Line; in one sense,
Black has gained a tempo due to
i..d3-e2, although it's not that easy) 1 2
l:.e l 4Jg6 1 3 i. f l 4Jg4! 1 4 4Jc4 ( 1 4
h3 4Jxf2 ! I S �xf2 .i.d4+ 1 6 'ue3 fS)
l 4 . . . .i.d4 I S 4Je3 'i!Vf6 = 16 4Jxg4
i.. xg4 17 'i!Vc2 4Jf4 1 8 i..e3 i.eS ! 1 9 f3
i.d7 20 �h l 4Jg6 2 1 'i!Vd2 h6 22 .i.e2
61
bS ! 23 a4 (23 4JxbS i.xbS 24 i.xbS
i.xb2 +) 23 ...bxa4 24 4Jxa4 i.xa4 2S
�xa4 l:tfb8 + with the customary
queenside pressure, Barstatis-Etruk,
Leningrad 1 962.
B
9. .4Jb6
.
The almost unknown 9 ...'iVe7 ! ? may
be another acceptable solution here.
Here's a short overview:
a) 10 'iVe2 is one of the few moves
actually tried here: 1 0... 0-0 1 1 i.f4
4Je8 1 2 g3 b6 1 3 a4 ( 1 3 4Jb5 4JeS =)
1 3 ... i.a6 ( 1 3 ... gS ! ? 1 4 i.d2 i.a6) 1 4
4JbS 4JeS I S i.xeS i.xeS 1 6 f4 ( 1 6
4JxeS 'i!VxeS 1 7 f4 'ViiIe 7) 1 6. . .i.d4 ! ?
( 1 6 ...i.. xbS 1 7 axbS i.d4 1 8 eS l:td8
1 9 i.g2 dxeS 20 fxeS f6 ! 2 1 e6 4Jd6 22
4Jxd6 �xd6 followed by . . .fS is com­
plex) 17 4Jxd4? ! ( 1 7 i.g2 i.xbS 1 8
axb5 4Jc7) 1 7 ... cxd4 1 8 �d 1 ? ? i.xc4
0- 1 Andruet - Santo-Roman, Epinal
1 986.
b) 10 i.d3 ! ? looks natural, but
shouldn't intimidate anyone: 1 0 ...0-0
1 1 0-0 4JeS ! 12 4JxeS 'ViiIxeS 1 3 'i!Vf3
( 1 3 h3 a6 1 4 a4 'ue8 =; 1 3 f4 'ViiId4+ 1 4
�h l i..g4 !) 1 3 . . . 'i!Ve7 (or 1 3. .. 4Jd7 1 4
62 THE GAMBIT GUIDE TO THE MODERN BENONI
..tf4 �e7) 1 4 ..tf4 ( 1 4 ..tgS h6 I S ..th4
gS 1 6 ..tg3 ..tg4 1 7 �e3 4JhS = )
1 4 ...4JhS IS ..te3 fS ! ? ( I S ... a6 16 a4
�4 also looks effective) 1 6 1:tae l
fxe4 1 7 'iWxe4 �xe4 1 8 4Jxe4 ..tfS =.
c) 1 0 ..te2 is the move Schneider
gives as favouring White: 1 0...0-0 1 1
f3, and now 1 l . ..4JeS 1 2 4Je3 a6 1 3 a4
1:tb8 1 4 0-0 was probably somewhat
better for White in Gorelov-Khasin,
Moscow 1 978, but a direct solution is
1 l ...4Je8 ! 12 0-0 fS 1 3 exfS 1:txfS 14
f4 ( 1 4 ..td3 can be met by 14 ... ..td4+
I S 'ith 1 l::t hS ! threatening . . .1:txh2+,
or even 1 4 .....txc3 I S bxc3 1:txdS)
14 ...1:tf8 I S ..tg4 4Jb6 =.
d) The untried 10 ..tf4 is critical,
since the ending after 1 O .. .'�Jxe4 1 1
4Jxe4 'iWxe4+ 1 2 'iWe2 �xe2+ 1 3 ..txe2
0-0 is double-edged. I think that Black
holds his own after 1 4 O-O-O! ( 1 4
..txd6 :e8 =) 1 4. . .4Jb6 ! ? ( 14. . ...td4 ! ?)
and now:
d l ) I S 4Jxb6? axb6 1 6 ..txd6 1Id8
17 ..tc7 ..tfS ! 1 8 g4 ( 1 8 ..txd8? fails to
1 8 . . ...th6+ 1 9 1:td2 1:txa2) 1 8 ... ..th6+
1 9 f4 :d7 +.
d2) IS ..txd6 ':d8 and now 1 6
..txcS? 4JxdS looks equal at first, but
it's very hard to defend White's king
in the face of .....te6 and ...:tac8; e.g.,
1 7 ..tf3 i.e6 1 8 4Jd6 1:tac8! 1 9 4Jxb7
1:txcS+ 20 4JxcS 1:tc8 +. Therefore
White should play 1 6 ..tc7, when
1 6 ...i.h6+ intending 17 'itbl 1:txdS or
17 f4 :e8 is equal, or Black might keep
more play on the board by 1 6 ... 1:te8 =.
d3) I S 4Jxd6 1:td8 ! ? (leading to dy­
namic play; also possible is simply
I S . . . ..td4) 1 6 .tf3 (this move looks
very strong, but Black has resources;
instead 16 4:Jxc8 1:taxc8 17 d6 ..td4 is
equal; e.g., 18 ..tg3 { or 18 ':'d2 h6 ! 1 9
..td3 4:JdS 20 ..tg3 4:Jf6 2 1 1:t e l 4:Je8,
leading to equality } 1 8 ... a6 19 ':he 1
�d7 20 ..tf3 l:tcd8 2 1 ..th4 ! ? f6 22 ..tg3
'itf8 ! = intending . . . 4:Jc8 or . . .4:Jc4j
16 .. Jhd6! (the point) 17 ..txd6 4:Jc4
1 8 ..tf4 ! ( 1 8 ..txcS?! i.fS ! 1 9 1:tdfl
':c8 ! +) 1 8 ... ..tfS ! ; e.g., 19 .l:Ide l 4:Jxb2
20 ..te2 4:Ja4 (20...c4 ! ? 2 1 d6 4:Ja4 22
..tf3 4:JcS) 2 1 d6 ..tb2+ 22 'itd l 4:Jc3+
23 'itd2 4:JdS with good prospects.
If 9 ...'Wie7 holds up, then it would
reinforce the overall ineffectiveness of
White's set-up.
We now return to 9 ... 4:Jb6 (D):
10 4:Je3
This poses more problems than 1 0
..tgS; for example, 1O...4:Jxc4 1 1 ..txc4
h6 1 2 .th4 0-0 1 3 0-0 a6 1 4 a4 ..td7 ! ?
I S 1:tel 'Wib6! 1 6 'iVc2 ( 1 6 l::tb l 4:JhS =)
1 6. . .gS 17 ..tg3 4:JhS 1 8 as 'Wic7 1 9 eS !
dxeS ( l 9 . . . 4:Jxg3 20 exd6 �xd6 2 1
hxg3 fS =) 20 d6 'iVxd6 2 1 1:tad l 'ilVc6
22 ..txeS ..txeS 23 l::txeS 1:tae8 =
I.Ivanov-Kudrin, New York 1 983.
10...0-0 n ..td3
 THE KNIGHT'S TOUR VARIATION
Or:
a) 1 1 a4 is well met by 1 1 ...i..d7 ! ?
(or 1 1 ....:te8 1 2 a5 tiJbd7 1 3 i..d3 tiJe5
= ) 12 a5 tiJc8 1 3 i..d3 l:te8 1 4 0-0 b5
15 axb6 tiJxb6 =.
b) 1 1 i..e2 'iie7 ! 12 'iVc2 i..d7 1 3
a4 llae8 and now, instead of 1 4 a5 '?"
as played in Anishchenko-Mochalov,
Minsk 1 974, Kapengut suggests 1 4
..Itf3 h 5 ( 1 4 . . .tiJxe4 ! ? 1 5 tiJxe4 f5) 1 5
h3 tiJh7 +.
1l ...:e8
The normal move, but the alterna­
tives aren't bad, and line 'b' is particu­
larly intriguing:
a) l 1 . ..tiJfd7 1 2 0-0 ( 1 2 f4 'ii'h4+
13 g3 'ike7 1 4 0-0 i..d4 and now 1 5
'it>g2 tiJf6 or 1 5 :el tiJf6 1 6 tiJb5 i..g4
17 'iVd2 i..f3 1 8 tiJxd4 cxd4 is equal)
12 ... tiJe5 1 3 i..e2 f5 14 exf5 ! ? ( 1 4 f4
tiJf7 1 5 exf5 gxf5 and now, instead of
16 �h l i.. d7, Reeh-Bilek, Kecskemet
1990, I think that White retains a small
edge after 1 6 i..d3 ! tiJh6 1 7 'iVh5 'iVf6
;1;) 14... gxf5 1 5 f4 tiJg6 1 6 tiJc4 tiJxc4
17 i..xc4 a6 1 8 a4 'iVf6 = Valette-Lecu­
yer, Chartres 1 990 (or 1 8 ...11e8 =) .
b) 1 1 . . .tiJh5 is underrated:
b l ) 1 2 g3 i..h 3 ! ? 1 3 a4 (Kapengut
queries Black's last move owing to 1 3
1i'f3 intending g4, but then 1 3 ... tiJd7
looks good; e.g., 1 4 g4? tiJe5, 1 4 tiJc4
f5 ! or 1 4 'iVe2 i..x c3+! 1 5 bxc3 'iVf6
intending ... tiJe5) 1 3 . . .:e8 1 4 a5 tiJd7
= 1 6 :bl tiJh3+ ! 1 7 �h l 'iixf3 1 8 gxf3
b2) After 1 2 0-0, the famous en­
counter Nimzowitsch-Marshall, New
York 1 927 continued 1 2. . . i..e5 1 3 a4
ttJf4 14 a5 tiJd7 1 5 tiJc4 with White
havi ng a slight positional edge (al­
though Hebert shows that even the
63
course of the game was OK for Black).
I think that 12 . . .tiJf4 ! (D) is better:
w
b2 1 ) 1 3 a4 tiJxd3 14 'iVxd3 :e8 1 5
a5 tiJd7 1 6 f4 i..d4 and now 1 7 �hl
tiJf6 is old analysis by Nunn, who sug­
gested that White was still better, but
Black is threatening ...i..xc3 and has a
great game; for example, 1 8 tiJed l
tiJxe4 ! 1 9 tiJxe4 i..f5 20 l:tel 'iVe7 +.
Instead, 17 tiJb5 is strongly met by
1 7. . . i..xe3+ ! 1 8 i..xe3 tiJf6, in view of
1 9 tiJc3? ! tiJxe4 20 tiJxe4 i..f5 2 1 lta4
'ii'e7 -+.
b22) 1 3 �h l tiJxd3 14 'iVxd3 f5 ! ?
1 5 exf5 i..xf5 1 6 tiJxf5 l:txf5, contem­
plating ... i..x c3, leaves Black at least
equal.
b23) 13 i..c2 (I think that this makes
the most sense) 1 3 .. Jle8 1 4 'iVf3 ( 14
i..d2 'iih4 ! ?) 14 . . . i..xc3 15 bxc3 'iVf6
tiJf4 ( 1 8 . . .tiJg5 ! ?) 1 9 tiJf5 i.. xf5 20
i.. xf4 tiJxd5 ! 2 1 i..x d6! tiJxc3 22 exf5
tiJxbl 23 :xbl b6 gives Black ade­
quate chances due to White's horrid
pawn-structure.
We now return to 1 1 .. .:e8 (D):
64 THE GAMBIT GUIDE TO THE MODERN BENONI
w worse.
12 0-0 Ji.d7!?
This is an experimental deviation
from theory. According to Schneider,
1 2 . . .c4 1 3 Ji.c2 Ji.d7 leads to equal
play. Perhaps true, but it's a little awk­
ward and I don't like to commit to an
early ...c4 in positions like this.
Several books feature Kapengut's
12 ... liJbd7 13 a4 a6:
a) 14 f4 c4 ! IS liJxc4 ( 1 S i.c2 ! ?
liJcs 1 6 'ii'f3 liJb3 ! is unclear accord­
ing to Kapengut; things look equal af­
ter 1 7 Ji.xb3 cxb3 intending ...'fIe7 or
..."iVb6) I S . . . liJcs is cited as 'unclear'
according to Nikolaev. At first I was
sceptical, but I think that Black has
full compensation: 16 'ii'f3 ( 1 6 'iVc2
liJg4 1 7 g3 fS ; 1 6 .l::.e l ? liJg4 1 7 g3?
Ji.d4+ 18 Ji.e3 liJxe3 19 liJxe3 'lib6
-+; 1 6 eS dxeS 1 7 fxeS liJxdS { Kap­
engut } 1 8 Ji.xg6?! Ji.e6 ! 1 9 Ji.xf7+
Ji.xf7 20 l:!.xf7 liJxc3 2 1 l:txg7+ �xg7
22 'ii'g4+ �h8 23 bxc3 l:tg8 ultimately
favours Black) 16 ...liJg4! 17 Ji.c2 Ji.d4+
1 8 liJe3 fS ! +.
b) Perhaps 1 4 as is better; then
Kapengut gives 14 . . . liJeS I S Ji.e2! gS
16 h3 liJg6 ;1;, although after 17 'lic2
liJf4 1 8 liJc4 liJxe2+ 1 9 �xe2 h6 (20
f4 g4), I'm not sure that Black is really
13 a4 1:tb8 14 aS liJc8
We shall encounter this type of
position elsewhere. Black's pieces all
exert influence, and he is ready for ac­
tion on the queenside. For example. . .
15 f4 'iVc7 16 lIel
1 6 "iVf3 bS 1 7 axb6 axb6 (the alter­
native 1 7 ...liJxb6 1 8 Ji.d2 c4 1 9 Ji.c2
'i!i'cs 20 �h l is playable but less at­
tractive) 1 8 Ji.d2 bS 1 9 �fc l c4 20
Ji.bl 'ii'cs with excellent activity.
16 bS
•.•
More direct than 1 6 ... a6 ! ? with the
idea ...liJa7.
17 axb6 axb6
The game is equal. Black will play
...bS, since White cannot occupy bS
without ceding the e4-pawn.
The nice thing about 7 . . .liJbd7 is
that Black has alternatives along the
way. Even the obscure 9 ...'i!i'e7 ! ? looks
interesting.
B)
7...Ji.g7 8 liJc4 0-0 (D)
w
 THE KNIGHT'S TOUR VARIATION
This is the traditional main line.
White will try to pressure the d-pawn.
He begins with:
Bl: 9 �f4 65
B2: 9 �gS 67
Or:
a) 9 e4 allows an immediate cen­
tral counterattack: 9 . . .tte8 (or 9 ... b5
10 lLixb5 lLixe4) 1 0 �d3 and now:
a l ) 1 O. . . b5? ! should be met not by
1 1 lLid2? a6 1 2 0-0 c4 1 3 �c2 lLibd7 +
J unquera-J .Lopez, Spanish Cht 1 990,
but 1 1 lLixb5 ! lLixe4 1 2 0-0 a6 1 3
lLiba3 lLif6 1 4 �f4 ;t.
a2) 10 . . .lLixe4 ! 1 1 lLixe4 f5 is best;
e.g., 12 �f4 fxe4 1 3 �c2 �f6 ! + with
the point 14 lLixd6 l:td8 1 5 lLixc8 �xf4
-+.
b) 9 g3 has recently become the
most popular move, probably because
White doesn't get much from the main
lines. The idea is to get a form of the
Fianchetto Systems of Chapter 6,
while bypassing some of the problems
there. However, the early positioning
of the knight on c4 can be committal,
and attack by . . .b6 and . . .�a6 tends to
disrupt White's plans. In fact, I like the
direct way Black played in Sadler­
Ward, British Ch (Nottingham) 1 996:
9 . .. b6 (similar . . .b6-based lines are
9 . . . lLia6 1 0 �g2 lLic7 1 1 0-0 lLife8 1 2
a4 b6 1 3 l:.el .i.a6 1 4 lLia3 f5 = Buh­
mann-Loetscher, Oropesa del Mar
1 999, and 9 .. :f1e7 1 0 �g2 lLibd7 1 1
�f4 lLie8 1 2 �d2 b6 1 3 a4 .i.a6 14
LtJb5 lLie5 = Delemarre-Reinderman,
Wijk aan Zee 2000, when 1 5 �e3 f6
16 h4 .i.xb5 1 7 axb5 lLic7 = would be
a sample line) 1 0 a4 lLie8 1 1 .i.g2 lLid7
65
1 2 0-0 �a6 1 3 lLia3 ( 1 3 lLib5 lLie5 14
lLixe5 �xe5 = 1 5 f4? .i.g7 1 6 e4?
�xb5 17 axb5 lLic7 + with the idea
that 1 8 �b3 is answered by 1 8. . .�d7)
1 3 . . .f5 (always a useful space-gaining
move in positions where lLie6 is not on
the cards) 14 �d2 lLie5 = 15 �c l lLif6
1 6 �h6 'iWd7 1 7 :a2 .i.xh6! 1 8 'iVxh6
lLih5 (threatening .. .f4 and . . .lLif7) 19
�c 1 f4 ! 20 lLiab5 h6 ! ? (20. . . �b7 ! ) 21
b3 :f7 22 lLie4 ! �xb5 23 axb5 :d8
(threatening .. :ilVxb5) 24 �c3 lLif6 25
lLixf6+ :xf6 26 IHa l ! ? (26 gxf4 l:txf4
27 �h3 ! =) 26 .. :�·xb5 27 gxf4 l:txf4
28 e3? (28 'iVe3 ! =) 28 . . . .:tb4 29 f4
':xb3 30 'ii'c2 lLid3 3 1 l:txa7 c4 32 �c7
�e8 ! -+.
81)
9 �f4 (D)
9 lLie8
••.
I think that a very close examina­
tion of the pawn sacrifices initiated by
9 . . . b6, 9 ...lLia6 and 9 . . .lLibd7 1eads one
to the conclusion that none of them
are quite sound (ECO, NCO, and most
alternative sources to the contrary).
66 THE GAMBIT GUIDE TO THE MODERN BENONI
Schneider gives the very best coverage
of these moves, although his presenta­
tion of the same pawn sacrifices after
9 i.gS h6 1 0 i.f4 (which are also du­
bious) is inconsistent. In any case, I
don't want to suggest such a committal
course unless I have confidence in it.
But there is nothing wrong with
9 ...ttJe8, which has a 30 or 40 year
history of soundness. Black prepares
...b6 and ... i.a6, as well as the space-
gaining . . .fS. It also turns out that
White's direct attempts to win the d­
pawn tend to fail in dramatic fashion.
10 'ikd2
Or:
a) 10 ttJe4? bS 1 1 ttJcxd6 ttJxd6 1 2
i.xd6 l:.e8 -+.
b) 10 e4 fS ! - Nunn.
c) 10 ttJbS? ! i.d7 ! 1 1 ttJbxd6 bS
12 ttJxe8 ( 1 2 ttJd2 'ikf6 1 3 ttJxe8 ':xe8
14 i.xb8 ':axb8 +) 1 2 ...i.xe8 1 3 ttJeS
'iWd6 14 ttJd3 'ikxdS + Psakhis.
d) 1 0 e3 is solid but a little slow:
1 0...gS ! (or 10 ...ttJa6 1 1 1.e2 ttJac7 1 2
a4 b6 1 3 0-0 i.a6 1 4 'ikb3 { 14 e4 fS !
Barsov-Galje, Haarlem 1 998 } 14 ... gS !
I S 1.g3 fS =) 1 1 i.g3 fS 1 2 'ikd2 'ike7
( 1 2. . .1.xc3 ! ? 1 3 bxc3 bS 1 4 ttJb2 'ike7
is unclear) 1 3 f4 ( 1 3 1.e2 ttJd7 1 4 0-0
{ 14 h4 g4 = } 14 ... i.xc3 I S bxc3 ttJdf6
1 6 f3 b6) 1 3 ...gxf4 1 4 1.xf4 ttJd7 I S
1.d3 ttJeS 1 6 0-0 i.d7 ( '+' Schneider;
maybe just = ) 1 7 e4? ( 1 7 ttJxeS dxeS
1 8 i.g3 ttJd6 =) 17 ...ttJxd3 1 8 'iWxd3
bS 1 9 ttJd2 b4 20 ttJe2 fxe4 2 1 ttJxe4
1.fS 22 ttJ2g3 i.g6 and Black has a
slight advantage, Renaze-Bouaziz,
Cannes 1 997.
10...b6 (D)
11 a4
w
1 1 e3 1.a6 1 2 a4 transposes to the
main line. Again, direct attacks on the
d-pawn don't achieve much:
a) 1 1 ttJbS 1.a6 1 2 a4 ( 1 2 ttJbxd6
ttJxd6 1 3 ttJxd6 transposes to line 'b')
12 ...1.xbS 13 axbS ttJd7 ! ? 1 4 ttJxd6
( 1 4 ':a6 ttJdf6) 14 ...ttJxd6 I S 1.xd6
l:.e8. Here Black has very active pieces
and White is underdeveloped: 1 6 i.g3
( 1 6 e3 ttJf6 17 1.g3 ttJe4 1 8 'ikc2
ttJxg3 1 9 hxg3 'ikxdS +) 1 6 .. J�e4 ! ?
( 16 ... ttJf6 i s also promising; e.g., 1 7
:dl ttJe4 1 8 'ikc2 and now 1 8 ...1.d4!
or 1 8 ...hS) 17 e3 l:.b4 18 l:.a2 ttJf6 1 9
d6 ttJe4 20 'ikd3 (20 'ikc2 ttJxg3 2 1
hxg3 'ikxd6) 20...'ikd7, intending ...'ike6
or just ...l:.d8, gives Black more than
enough compensation.
b) 1 1 ttJe4 1.a6 1 2 ttJcxd6 ttJxd6
1 3 ttJxd6 ( 1 3 1.xd6 ':e8 14 1.xb8
':'xe4 +) 1 3 ... gS 14 1.g3 ( 1 4 ttJxf7 ! ?
lhf7 I S 1.xgS 'tWd6 1 6 g 3 and now,
instead of the greedy 1 6 ... 1.xb2? as in
Lorbek-Strbad, COIT. 1 992, Black keeps
a pleasant advantage with 1 6... 1.d4 ! )
1 4... fS I S e4 1.xfl 1 6 ':'xfl f4 1 7 ttJfS
fxg3 +. White has some but not enough
compensation for the piece.
 THE KNIGHT'S TOUR VARIATION 67

1l i.a6 12 e3
.•.
12 liJb5 i.xb5 1 3 axb5 f5 leads to
equality: 1 4 h4 ( 1 4 e3 1:1f7 1 5 i.e2 a6)
1 4... i.f6 1 5 e3 :f7 1 6 i.e2 a6 ! 1 7
bxa6 IHa7 =. A resource to remember.
12 f5
•..
This move allows Black to gain a
tempo in comparison to 1 2" .i.xc4 if
White plays 1 3 i.e2.
I S 'itfl ( 1 S 0-0 lIeS ! 1 9 'it'c2 liJe5 20
i.e2 liJbd7 = ) Borisenko-Boleslavsky,
USSR Ch (Moscow) 1 96 1 ; then
I S ...liJe5 ! 19 i.c2 liJf7 20 'it>g2 lIeS =
is best.
For a third option, 1 6 ... lIeS is solid.
1 7 i..g5
1 7 i.h2 is met by 1 7 ...'it'xh4.
17 'it'd7!
•.•

13 h4 The idea is to chase the bishop by

Or: ...h6 while keeping an eye on g4 and
a) 1 3 i.e2 'iWf6 (threatening ".g5 discouraging 0-0-0 in view of the a­
and ".f4) 14 i..g3 i.xc4 15 i.xc4 a6 pawn. 1 7 ...'it'eS has the same idea of
16 0-0 liJd7 = Osnos-Forintos, Lenin­ ... h6. Then a possible continuation is
grad-Budapest 1 962. IS 'it'c2 ! liJd7 1 9 i.e2 f4! 20 i.g4!
b) 1 3 i.g3 ? ! 'iVe7 ! (threatening fxe3 2 1 i.e6+ 'ithS 22 i.xe3 liJf4 23
.. .f4) 14 f4? ( 1 4 0-0-0 i..xc4 1 5 i.xc4 i.xf4 lIxf4, and Black answers both
a6 16 'itbl liJd7 1 7 llhel b5 ! I S axb5 24 h5 and 24 liJe4 with 24...liJfS ! .
ruc7 gives Black a serious attack) After the text-move ( 17 . . .'it'd7), a
14".i.xc3 ! ? ( 1 4... liJc7 ! +; 14... liJf6 1 5 likely continuation is I S 'it'c2 'it'f7 1 9
i..h4 11eS, Bumett-Nunn, Oxford 1972, liJe2 liJd7, when Black has n o worries.
and now White should try 1 6 liJb5) 1 5
bxc3 'iVe4 1 6 liJb2 i.xfl 1 7 lIxfl liJf6 82)
=t. 9 i.g5 (D)
13...i.xc4 14 i.xc4 a6 15 'it'e2
1 5 i.e2 liJd7 ('+' ECO) 1 6 e4
(better is 1 6 i.g5 ! , when 1 6 ...'it'c7 1 7
g4 i s double-edged and after 1 6 ...i.f6
17 i.h6, 1 7 . . .:f7 ! ? { with the point I S
h5 g5 ! } tries to avoid the repetition
17 . . . i.g7 1 S i.g5, etc.) 1 6 .. .fxe4 1 7 h5
ctJdf6 I S hxg6 hxg6 1 9 lIh3 'iWd7 20
0-0-0 b5 ! 2 1 axb5 axb5 22 i.xb5 'iWf5
23 i.e3 liJg4 + Farago-Velimirovic,
Amsterdam 1 976.
15 liJf6 16 i.d3
.••

';!;' - Kapengut's annotation, but I

doubt it.
16 liJh5!?
•.. This move is considerably more
Perhaps this is better than the previ­ popular than 9 i.f4, because the sim­
ously-played 16 ...lIa7 ! ? 17 g 3 ! ? liJg4 ple solution in the last section is not
68 THE GAMBIT GUIDE TO THE MODERN BENONI
available (see note 'a2' below). Left
alone, White might play 'iVd2-f4. From
Black's point of view, he can be happy
that he has quite a bit more leeway in
defence. I will recommend two con­
tinuations:
B21: 9 :ii'd7!? 68
..
B22: 9 'iVe7
... 70
Here are thoughts on two other
ideas:
a) I admit that I like the popular
9 ... h6 for White after 1 0 .lif4 since the
pawn sacrifices mentioned in Line B 1
still fall short, in my opinion.
a l ) In that context, I should men­
tion that 10 ...ttJa6 ! ? 1 1 'iVd2 b5, lead­
ing to complex play after 1 2 ttJxb5
ttJe4 in the often-quoted game Kap­
lun-Kapengut, Rostov 1 980, is well
answered by 12 ttJxd6 ! b4 ( 1 2...ttJh5
1 3 ttJdxb5; 1 2 . . . .lid7 1 3 e3 { or even
1 3 e4 } 1 3 ...b4 14 ttJce4) 1 3 ttJcb5 ( 1 3
ttJdl ! ?) 1 3 ....lid7 ( 1 3 ... g5 1 4 .i.g3 ttJh5
1 5 ttJxc8 ttJxg3 1 6 hxg3 "iWxc8 1 7 e3)
14 e4 ! .
a2) Thus, Black should probably
be content with 1 O. . .ttJe8, when the
difference between this and the last
section is 1 1 'iVc 1 ! g5 ( 1 l .. .�h7 is met
by 1 2 a4! with the idea ttJb5, and not
1 2 ttJb5 .lid7 ! , which I suggested in
my book, and gives Black good play)
12 .lid2, intending h4. Theory is di­
vided, but I think that this is very un­
comfortable for Black.
b) 9 ...ttJa6 may be underestimated:
10 e3 ttJc7 1 1 a4 b6 1 2 .lie2 .lia6 1 3
0-0 'iVd7 (D) and then:
b l ) 14 b3 'iVf5 ! ? ( 14 . . .ttJfe8 1 5
.:tc l f5 1 6 .lif3 .lib7 ! i s also fine) 1 5
W
.lif4 ttJcxd5 1 6 ttJxd5 'ilVxd5 = 1 7
.lixd6 :fe8 1 8 .lif3 'iVxd l 1 9 �fxdl
ttJe4 with balanced play.
b2) 14 h3 �fe8 1 5 �c 1 .lixc4 1 6
.lixc4, Portisch-Ionescu, Moscow OL
1 994, and now lonescu gives 16 ...ttJe4
=, which looks rather promising for
Black in view of lines like 1 7 ttJxe4
�he4 1 8 'iVc2 �ae8 1 9 f3 ( 1 9 .lif4 a6)
1 9 . . J:t4e5 20 .i.f4 1hd5 ! 2 1 .lixd5 (21
g4 a6 ! and ...b5) 2 l .. .ttJxd5 22 I:tfel
ttJxf4 23 exf4 .lid4+ 24 �f1 .lie3 ! +.
b3) 14 e4 ! ? (wasting a tempo seems
a bit strange, but White is not happy
with his restricted centre) 14 ... .lixc4
1 5 .i.xc4 a6 1 6 'iWf3 ( 1 6 'iVe2 ttJg4 1 7
h 3 ttJe5 1 8 .lib3 f5 1 9 exf5 gxf5 ! ? 20
f4 ttJg6 = Gil Capape-Suba, Ponfer­
rada 1 992) 1 6 ...ttJg4 1 7 .i.e2 f5 (or
17 ...ttJe5 1 8 'ilVg3 f6 1 9 .i.e3 f5 20 exf5
'ili'xf5) 1 8 exf5 gxf5 with equality,
Ehlvest-Ionescu, Moscow OL 1 994.
82 1 )
9 :iVd7! ? (D)
.•
I give this relatively untested solu­
tion for its ingenuity as well as its
merit. Black escapes the pin on his
 THE KNIGHT'S TOUR VARIATION
f6-knight, prepares ...bS, and has the
tactical ideas of ...�g4 or ...'iVfS at the
right point.
10 a4
White decides that his first priority
is to prevent ... bS . Others:
a) After 1 0 i.f4 ! ?, Hebert invests
some intelligent analysis on 1 O...ttJe8,
but l O ... bS ! (D) is more in the spirit of
this line:
I I ttJxd6 ttJhS 1 2 �d2 gS ! 1 3 i.eS !
�xeS (or 1 3 ... f6 ! ? 14 ttJxc8 fxeS I S
d6 { I S �xgS ttJf6 and one has to like
Black} IS . . .:xc8 16 'ili'dS+ 'it>h8 1 7
69
�xa8 ttJc6) 1 4 �xgS+ i.g7 I S ttJxc8
ttJf6 1 6 e4 ( 1 6 ttJxbS ttJe4 ! ) 1 6...�xc8
17 i.xbS ( 1 7 eS �g4) 17 .. .'ifb7 ! 1 8
0-0-0 ttJxe4 1 9 ttJxe4 'iifxbS 20 ttJf6+
'it>h8 2 1 �fS 'iifc4+ 22 'it>b I 1\th4
leaves Black at least equal, one sur­
prising line being 23 ttJg4? ttJd7 ! 24
�xd7? fS 2S ttJe3 'iib4 ! -+.
b) 1 0 e3 bS 1 1 i.xf6 i.xf6 1 2 ttJe4
i.g7 1 3 ttJcd2 ( 1 3 ttJcxd6 fS ! 1 4 i.xbS
'iifd 8 I S ttJxc8 fxe4 1 6 d6 'ifaS+ +)
1 3 ... i.b7 1 4 a4 a6 I S axbS axbS 1 6
Iha8 i.xa8 1 7 ttJf3 i.xb2 1 8 � b l ( 1 8
'ilfb3 i.g7 1 9 i.xbS i.xdS !) 1 8 ...i.xdS
1 9 i.xbS �xbS 20 ttJf6+ 'it>h8 2 1
ttJxdS �a5+ 22 ttJd2 i.g7 2 3 'ifc2 c4
with a winning position for Black,
N.Michaelsen-Lautier, Hamburg 1 986.
10 ttJa6
•••
Now that the b4-square is free.
11 e3
Or:
a) In comparison with the previous
note, 1 1 i.f4 ttJe8 appears less de­
manding for Black, who has ...ttJb4 in
reserve. Then after 12 ttJbS, 1 2 ...ttJb4 !
1 3 ttJbxd6 ttJxd6 14 i.xd6 i.xb2! is
fine, among others.
b) 1 1 �d2 ttJb4 1 2 f3 ! ? (Laren­
S.Wolff, corr. 1 990) and now 1 2 ...ttJe8!
would anticipate 1 3 e4 fS ! .
11 ttJb4 1 2 i.e2 �f5!
••.
One of the points of 9 .. .'iVd7: to hit
gS and c2 at the same time.
13 i.xf6 �xf6 14 0-0 �e7 15 'ili'd2
Black is equal after IS .i:tc 1 i.d7 or
l S ... i.fS .
After the text-move ( 1 S 'iVd2), the
game Popov-Romanishin, Stara Paz­
ova 1988 continued I S ...b6 ( 1 S ...i.fS ! ?)
16 ttJa2 ttJxa2 17 Iha2 i.a6 ! ? (better
70 THE GAMBIT GUIDE TO THE MODERN BENONI
is 1 7 ... .tb7 ! 1 8 as bS 1 9 a6 .txa6 20
l:lxa6 bxc4 with at least equality) 1 8
b3 l:tab8 1 9 l:lc 1 .tb7 20 .tf3 l:lfd8
with about equal chances ( ... .tc8 fol­
lows).
822)
9. .'fie7 (D)
. 'fih4 ! 2 1 l:lgl c4 ! with active play,
w 11 'fif4 l:ld8 (D)
This, our second repertoire move, is
one of the most frequent answers to 9
.tgS. Black refuses to create a weak­
ness by ... h6, and prepares moves like
. . .tLlbd7 and/or ...b6 and ... .ta6.
10 'fid2!
The most challenging move. 10 e3 is
not so ambitious: 1 O ... tLlbd7 ( l 0...b6
is a sound alternative) 1 1 .te2 tLleS
and then:
a) 1 2 tLld2 is slow, and well met by
12 ...h6 1 3 .th4 a6 14 a4 l:le8 I S 0-0
l:lb8 = (or IS ...'fic7 ; or even I S ....td7
1 6 h3 l:lab8, since 1 7 f4 is met by
1 7 ... tLld3 !).
b) 1 2 tLlxeS 1 2 ...'fixeS 1 3 .tf4 'fie7
1 4 0-0 .tfS I S l:le l a6 1 6 a4 l:lfb8 1 7
h 3 .td7 = intending ...b S , Gulko­
Wahls, Groningen 1 990.
10 b6
.•.
The controversial 10 ...tLlbd7 ! ? 1 1
'fif4 tLleS 1 2 tLle4 .tfS is too complex
to analyse here. The stem game went
1 3 tLlcxd6 .txe4 14 tLlxe4 tLled7 I S
tLlxf6+ .txf6 ! 1 6 .txf6 tLlxf6 1 7 0-0-0
l:lfd8 1 8 e4 tLlxe4 1 9 l:lel ! ? fS 20 g4?
N.Michaelsen-Palkovi, Eger 1 987.
However, this looks awfully specula­
tive at best.
12 0-0-0
Or:
a) 1 2 tLle4 'fixe4 1 3 'fixe4 tLlxe4 1 4
.txd8 b S (given b y Yusupov a s un­
clear) I S f3 bxc4 1 6 fxe4 .txb2 1 7
l:lbl c 3 1 8 .tgS tLld7 1 9 .te l tLlf6
leaves Black at least equal, with the
point 20 .txb2? :b8 - Schneider.
b) 1 2 f3 .ta6 1 3 e4 bS 1 4 tLld2 b4
I S tLld 1 h6 1 6 .txf6 ( 16 .txh6 tLlhS 1 7
'figS .tf6 1 8 �e3 .td4 =) 1 6. . ..txf6
17 'fie3 .txfl 1 8 'itxfl tLld7 = Eliet­
Kinsman, Toulouse 1 996.
c) I suggested 12 a4 in my book,
but 1 2 ... tLla6 ! disrupts White's play;
 THE KNIGHT'S TOUR VARIATION
e.g., 1 3 lLle4 ( 1 3 e3 lLlb4 14 �c 1 oltb7)
13 ...'i!Vxe4 14 'iixe4 lLlxe4 I S .txd8
ttJb4 1 6 l:tc 1 .ta6 ! ? ( 1 6 ...lLla2 17 l:ta1
ttJb4 1 8 �c 1 = ) 17 .tc7 lLlxdS 1 8 f3
ttJxc7 1 9 fxe4 oltxc4 20 ltxc4 .txb2
-+.
12....ta6 13 e4 .txc4 14 .txc4 a6
15 'iWh4
I S a4 lLlbd7 1 6 �he1 bS 1 7 axbS
ttJeS ! 1 8 oltfl axbS gives Black a clear
advantage.
15 ... lLlbd7 16 l:thel 1i'f8! (D)
w equality with an abundance of oppor­
17 �b1
Or:
a) 1 7 a4 bS ! 1 8 axbS axbS 1 9 .txbS
71
h6 20 .te3 l:tdb8 2 1 .tc6 l:ta6 with an
attack and positional pressure.
b) 1 7 eS dxeS 1 8 lLle4 lLlxe4 1 9
oltxd8 ':'xd8 20 l:txe4 bS with plenty of
compensation; e.g., 2 1 .tb3 c4 22
.tc2 lLlb6 +.
After the text-move ( 1 7 �bl ),
17 ...bS 1 8 oltfl h6 ! 1 9 .tc1 lLlb6 ! 20 f4
(20 .txh6 .txh6 2 1 �xf6 .tg7 22 'iff3
b4 23 lLle2 lLlc4) was Portisch-T.Hor­
vath, Hungarian Ch 1 984. Here Black
could have secured a clear advantage
by 20...b4! 2 1 eS dxeS 22 fxeS gS ! 23
'ifg3 bxc3 24 exf6 .txf6 :t.
Thus both 9 ...'ifd7 and 9 ...'ife7 yield
tunities to unbalance the play.
The Knight's Tour variation has
never been fully worked out, but as I
see it, both 7 ... lLlbd7 and the tradi­
tionaI 7 ....tg7 8 lLlc4 0-0 are fully suf­
ficient. One way to look at it is that
White has tipped his hand by playing
lLlf3-d2-c4 so early. This allows Black
to adjust comfortably with a set-up de­
signed to neutralize that manoeuvre.
Perhaps it's no surprise after all that
White has practically abandoned this
system in master play.
5 Pawn-Storm Systems

I d4 ttJf6 2 c4 cS 3 dS e6 4 ttJc3 exdS 5 Thus, material is divided into:

cxdS d6 6 e4 g6 7 f4 (D) A: 7 i..g7
••• 72
B: 7..:fIIe7 106

A)
B 7....i.g7
Now there are three very distinct
systems for White:
AI: 8 eS 72
A2: 8 ttJf3 75
A3: 8 .i.bS+ 82

AI)
8 eS
The Mikenas Attack, rarely seen
these days because White's centre be­
In this chapter, we look at a variety comes vulnerable.
of pawn-storm attacks by White, all 8...ttJfd7 (D)
beginning from the diagram position. Theory suggests that 8 . . . dxe5 9
White typically tries to overrun Black's fxe5 ttJfd7 may also suffice. 8 ...ttJfd7
position with e5, but the methods vary. is better established.
The most challenging of such attacks
is the 'Taimanov Benoni' with 7 ... .i.g7
8 .i.b5+. Benoni players will also
meet the Four Pawns Attack (8 ttJf3), a
favourite of many club and Swiss Sys­
tem players, and occasionally the
Mikenas Attack (8 e5). I will offer rep­
ertoires against each of these systems.
If one wants to take a chance and by­
pass all of them, I have also added a
section on a practically unknown vari­
ation (7 ..."ile7) that is risky, but might
interest the reader.
 PAWN-STORM SYSTEMS
9 lLlb5
Or 9 lLle4 dxeS 10 lLld6+, transpos­
ing. Others:
a) 9 e6? ! fxe6 10 dxe6 lLlb6 (the al­
ternative 1 O. . . .i.xc3+! 1 1 bxc3 lLlb6
may be even better; one must wonder
whether White's position is worth a
pawn) 1 1 lLle4 ! ( I I fS 0-0 +; 1 1 �b3
-rtIie7 1 2 .i.bS+ lLlc6 and . . . .i.xe6)
1 1 . . .0-0 (or l l ...dS ! ? 1 2 lLlxcs �e7 :+)
12 Wixd6 'ii'xd6 1 3 lLlxd6 .i.xe6 14
CLlxb7 CLla4 ! with a slight advantage
for Black.
b) 9 exd6 0-0 10 lLlf3 lLlf6 1 1 .i.e2
CLle8 (or l l . ..a6 1 2 a4 �xd6 =) 1 2 0-0
CLlxd6 = Psakhis.
c) 9 lLlf3 0-0 10 .i.e2 dxeS 1 1 0-0
CLla6 ( l 1 ...e4 ! ? 1 2 lLlxe4 lLlf6 1 3 lLlc3 !
CLle8 ! = Valsser; l l ...exf4 is also played)
12 .i.e3 ':'e8 1 3 'ii'd2 ( 1 3 fxeS lLlxeS :+)
1 3 ...e4! 1 4 lLlgS lLlf6 :+ Ki.Georgiev­
Semkov, Varna 1 982.
9...dxe5 10 lLld6+ �e7 ll lLlxcs+
Or:
a) 1 1 fxeS? ! lLlxeS 1 2 lLlxc8+ 'ii'xc8
1 3 d6+ �f8 14 lLlf3 'iVe6 I S lLlxeS
.Yi.xe5 1 6 .i.e2 �g7 left Black well on
top in Kavalek-Trapl, Prague 1 963.
b) I 1 lLlb5 (this is White's main al­
ternative to the text-move) 1 1 .. ..:.e8 ! ?
( l 1 .. .lLla6 has been quite successful
and may even be Black's most appeal­
ing option; e.g., 1 2 d6+ �f8 1 3 lLlf3
'i'f6 14 fxeS lLlxe5 1 5 .i.e2 lLlxf3+ 1 6
i.xf3 Wih4+ 1 7 �fl .i.d7 1 8 lLlc3 %le8
+ Maenner-Steinert, Bern 1 99 1 ) 1 2
d6+ �f8 13 lLlc7 exf4+ (D) and now:
b l ) 14 lLlxe8 Wixe8+ 1 5 .i.e2 lLle5
16 .i.xf4 lLlbc6 (with a strong initia­
tive) 1 7 lLlh3? .i.xh3 1 8 gxh3 lLlf3+ 1 9
'.t> f2 WVe4 20 .i.xf3 'iWxf4 2 1 �g2 lLld4
73
W
22 lk l lLlfS 0- 1 Smirnov-Kapengut,
Minsk 1 979.
b2) 14 .i.e2 and now:
b2 1) Black can consider 14 ...�h4+,
when after 1 5 �d2?, I S . . . f3 ! looks
strong; e.g., 16 lLlxf3 'iWb4+ 1 7 �c2
lLle5 1 8 lLlxeS ':'xeS :+ 1 9 d7? .i.xd7 20
'ii'd6+ �g8 2 1 .i.d3 ]:tf5 -+ Culli­
nane-Denman, British Ch (Eastboume)
1 973. 1 5 �fl is correct, when I think
that 1 5 . . . .i.d4 ( 1 5 . . .lLlb6 ! ?) 1 6 'ii'e 1
'iWxel + 17 �xel lLlb6 is interesting;
e.g., 1 8 lLlxe8 'iitxe8 1 9 lLlf3 lLlc6! or
1 8 i.xf4 .i.d7 ! 1 9 lLlxa8 ( 1 9 .i.h6+
'iitg 8 20 lLlxe8 .i.xe8 2 1 ':'bl iLc6 ! ? 22
lLlf3 .i.e4 23 ':'c 1 .i.xb2 24 l:lxc5 .i.a3
25 ':'e5 .i.c6 = ) 1 9 . . .lLlxa8 20 �fl
lLlb6 2 1 ':'c 1 lLld5 ! 22 .i.h6+ �g8 23
lLlf3 lLle3+ 24 iLxe3 .i.xe3 25 ':'d 1
lLlc6 =. White's king position is a
problem in these lines, but it wouldn't
be surprising if he still has a small
edge.
b22) Perhaps objectively best is
the older 14 ... lLlc6, which is active and
dynamically balanced. The main line
goes 1 5 lLlxe8 ( 1 5 lLlxa8 ! ? is men­
tioned by Nunn, but I haven't seen it
74 THE GAMBIT GUIDE TO THE MODERN BENONI
played or analysed) I S . . .'iWxe8 1 6 ttJf3
ttJd4 1 7 ttJxd4 i.xd4 1 8 i.xf4 ttJeS,
Taylor-Donnelly, corr. 1 990, when
Black has a lot of compensation (VaYs­
ser gives 1 9 'iWd2 i.fS 20 0-0-0 'iWa4 !).
In any case, the earlier possibility
l l . ..ttJa6 looks attractive, so 1 1 ttJbS
shouldn't overly worry Black.
1l :ii'xc8 12 ttJf3
..
1 2 d6+?! �f8 ( 1 2 . . . �d8 1 3 ttJf3
exf4 14 ttJgS ttJeS favoured Black in
K.Schneider-Schima, COIT. 1 989) 1 3
ttJf3 ( 1 3 'iWb3 e4! ) 1 3. . .e4 1 4 ttJgS h6
I S ttJxf7 ( 1 S ttJxe4 'iWe8 ! 1 6 'iWe2 ttJc6
+ Partos-Holm, Skopje OL 1 972)
IS ... �xf7 16 i.c4+ 'ifi>f8 1 7 fS, and at
this point Black has three very strong
moves: 1 7 . . . gS, 1 7 . . . i.d4! 1 8 fxg6
ttJeS (intending to answer 1 9 l:tfl +
with 1 9 ...'ifi>g7), and 1 7 ...ttJc6 1 8 fxg6
ttJdeS ! . Schneider continues 1 9 0-0+
�e8 20 i.f7+ �d8 2 1 'ildS, and apart
from his 2 l . ..'iWg4, 2 l .. .'ild7 22 i.e6
'ile8 + is simple enough.
12 e4!? (D)
•••
Povah's move, still not well investi­
gated and therefore with considerable
surprise value. On the positive side,
. . .e4 frees the monster on g7 and gains
a tempo. Just as importantly, it tempo­
rarily keeps both the e- and f-files
closed, affording some much-needed
protection to Black's king. On the neg­
ative side, Black fails to develop, and
he can no longer get his rook to e8 in
view of 1 2 ...e4 1 3 ttJgS l:te8?? 14 d6+
�f8 IS 'iVdS.
12 ...e4 might be the best method of
playing for an advantage, because
1 2 .. J�e8 ' only' equalizes in some
lines. On the other hand, 12 . . .l:te8 has
been heavily analysed and proven in
practice; the interested reader should
consult ECO or VaYsser.
w
13 ttJg5
Or:
a) 1 3 d6+ �f8 1 4 ttJgS at best
transposes after 14 . . .ttJb6.
b) 1 3 ttJeS ttJxeS 14 d6+ �f8 I S
fxeS, and now I S . . .'iVe6 !, among oth­
ers.
13 ttJb6! 14 d6+
•••
Nunn analyses these alternatives:
a) 14 i.e2 i.d4 ! .
b) 1 4 'iVb3 'iVfS I S d6+ �f8 1 6
i.c4 ( 1 6 g 3 i s well met b y 1 6 ... ttJc6 1 7
i.h3 'ii'dS ; Nunn gives 1 6 g4 'ild7,
when 17 i.e3 i.d4 1 8 i.bS ttJc6 1 9
i.xc6 bxc6 20 i.xd4 cxd4 2 1 'iWb4
ttJdS ! 22 'ilxd4 f6 is surprisingly good
for Black) 1 6...h6 1 7 ttJxf7 ttJxc4 1 8
'iWxb7 'iWxf7 1 9 'iWxa8 'iWe8 20 'iWxa7
i.d4 "with a strong initiative" - Nunn.
c) 14 ttJxe4 l:td8 gives Black good
play. A silly example continued I S
i.e3 �f8 ! 1 6 i.xcS+ 'ifi>g8 + 17 d6??
l:te8 -+ Crippa-Pontecorvo, COIT. 1 987.
14...�f8 15 a4
 PAWN-STORM SYSTEMS
Or I S i.c4 t'Llxc4 16 'iVdS t'Llxd6 17
'li'xd6+ �g8 18 0-0 i.d4+ 1 9 '>t>h 1
'li'fS ! + Kerr-Povah, London 1 976.
After the text-move ( 1 S a4), Kooi­
man-Povah, London 1 976 continued
IS . . .h6 1 6 as hxgS 17 axb6 a6 18 �dS
and now 1 8 .. :iVc6 with a complex po­
sition; instead, 1 8 ...'iVd7 ! followed by
. . . t'Llc6 must favour Black; for exam­
ple, 19 fxgS t'Llc6 20 i.bS i.d4 intend­
ing . . .cJi;g7 or ... 'iVg4.
A2)
8 t'Llf3 0-0 9 i.e2 (D)
I have played the extremely irregu­
lar 9 i.d3 a few times for its surprise
value, but there are several good an­
swers, the easiest being 9 ...i.g4 fol­
lowed by ...t'Llbd7 and restraint of the
e-pawn; for example, 10 0-0 t'Llbd7 1 1
h3 i.xf3 1 2 1'Vxf3 a6 1 3 a4 l:!.b8 (or
1 3 .. :it'aS) 14 as t'Lle8, as in our main
line below, intending ...t'Llc7 and ... bS .
9 . . .'iVb6 ! ? is also fascinating and al­
most unexplored. The main point is
that the natural 10 t'Lld2 t'Llg4 1 1 t'Llc4
\i'd8 ! is remarkably difficult to coun­
ter.
75
9 i.e2 introduces the main line of
the Four Pawns Attack. Since this
mostly arises via a King's Indian De­
fence, I will follow a policy of sug­
gesting two interesting variations, but
ones which don't require quite the de­
tail that I supply for true Benoni sys­
tems:
A21 : 9 i.g4
••• 76
A22: 9 t'Llbd7 79
.•.
I think that the second option is ex­
tremely underrated, and hope that you
at least consider playing it.
For the record, after years of looking
at it, I'm also convinced that 9 ...l:te8 is
absolutely sound, and even slightly
advantageous in some of the best­
known main lines. However, its com­
plete analysis would require too much
of this book, so for those interested, I
refer you to the books in the Bibliogra­
phy by Valsser and Schneider.
In addition to Line A22, another
knight move that is interesting and
perhaps underestimated is 9 . . .t'Lla6 ! ?:
a) There are few good examples,
but White's theoretical answer used to
be 10 eS dxeS 1 1 fxeS t'Llg4 1 2 i.f4 ( l 2
i.gS 'iVb6 1 3 t'Lla4 'iVb4+ 1 4 i.d2 'iVe4
I S t'Llc3 'iVfS is hard to assess, but
looks OK for Black) 1 2 ...l:!.e8 1 3 e6
fxe6 14 d6, and now instead of the pre­
viously played 14 ... i.d7, 14. . :tlt'b6! is
a clear improvement: IS 'iWb3 ( I S 'iVd2
is met by I S ... eS !, when 1 6 i.c4+?
i.e6 1 7 d7 exf4 gives Black a winning
position, while 1 6 i.bS l:te6! still fa­
vours Black) I s .. :iWxb3 16 axb3 t'Llb4
and Black has a distinct advantage,
Urbanek-Pfeifer, corr. 1 999.
76 THE GAMBIT GUIDE TO THE MODERN BENONI

b) So perhaps White should be

content with 1 0 0-0 lbc7; e.g., 1 1 eS t ?
( 1 1 a4 l:te8 = ; 1 1 l:!.e l lbd7 = , when 1 2
�e3 l:!.e8 threatens . . .i.xc3 and makes
defence of e4 awkward in view of 1 3
�d3 b6 or 1 3 lbd2? i.xc3 14 bxc3
lbxdS) l l . . .dxeS 12 d6 lbe6 1 3 fxeS
lbd7 =.
9 ... lba6 definitely deserves more
tests.

A2 1 )
9 i.g4
•••

This is a popular simplifying idea
that reduces the threat of eS by trying
to eliminate the f3-knight. If White
plays lbd2 and allows ... i.xe2, it will
be equally difficult to advance in the
centre. I will recommend this system
with an irregular twist on move 1 1 .
10 0-0
With 1 0 h3 i.xf3 1 1 �xf3 lbbd7 1 2
0-0, w e transpose to note 'd' to
�h8 20 l:!dl 'it'aS '+ Vai'sser) 1 9 1:I.fl
( 1 9 'iVh3 lbf8 20 l:td l �d4 ! ? '+; 19
l:.dl ! lbeS 20 'it'e4 'it'd6 ! '+) 19 ... lbeS
20 'iVe4 'iid6 21 i.gS lbc6 (Kapengut
suggests 2 1 . . Jle8) 22 lbf6+ (22 i.f6
:f8) 22 . . .�h8 23 �h l 'iVd4, Schoen­
A.V.lvanov, Bie1 1 990. Kapengut as­
sesses this as equal, but I see no real
compensation.
10 lbbd7 (D)
.••

White's 1 1th move. There is some point in 1O . . . i.xf3 1 1

I hate to pass by 10 eS lightly, since �xf3 ; for example, l l . ..lbfd7 ! ? is in­
it has a small, dedicated following, but teresting in that case. But 1O ... lbbd7 is
it has always been OK for Black. A more logical, saving a tempo if White
sample from the main line: 1 O. . ..i.xf3 plays h3.
1 1 i.xf3 dxeS 1 2 fxeS lbfd7 1 3 e6
lbeS (D) and now:
a) 14 exf7+ llxf7 I S 0-0 lbbd7 =
(or I s . . .lbxf3+); for example, 1 6 i.e2 w
Ihfl + 17 'iVxfl 'it'b6 = Reinemer­
Kuzmin, Oberwart 1 995.
b) 14 0-0 fxe6 IS i.e3 ! ( I S .i.e4
l:txfl + 1 6 'iVxfl �h8 ! ? 17 dxe6 can be
answered by 1 7 .. :i!Vd4+ 1 8 �h l lbg4
= or 1 7 . . . lbbc6 - Va'isser) I s ... lbxf3+

( I S ... lba6 ! ? is also OK) 1 6 1hf3 l:!xf3

1 7 �xf3 exdS 1 8 lbxdS lbd7 (more
straightforward is 1 8 . . . lbc6! 19 .i.xcs
 PAWN-STORM SYSTEMS
11 :el
Probably the most demanding move.
Other moves have been popular in the
past (especially 1 1 h3), but they seem
to have fallen out of favour:
a) 1 1 'iVc2 tDe8 1 2 i.d2 a6 1 3 a4
l:tc8 = 14 i.c4 ! ? i.d4+ 1 5 'it>hl tDef6 ! ?
16 l:.ael ne8 1 7 tD g l tDh5 = Akop­
yan-Rey, San Francisco 2000. The po­
sition is in a kind of dynamic balance.
b) 1 1 tDd2 i.xe2 1 2 �xe2 J:.e8 1 3
�f3 ( 1 3 tDc4 tDb6) 1 3 . . .'iVb6 ! ? (an
original idea; 1 3 ...J:.c8 14 'it>h l c4 1 5
g4 h6 ! = i s the conventional line, while
Blokh proposes the simple 1 3 . . .'ii'e7
14 l:te l tDb6! 1 5 J:.e2 c4) 1 4 tDc4 �a6 !
15 tDa3 (versus ...b5) 1 5 . . .c4 16 .ie3
tDxe4 ! 17 tDxe4 i.xb2 18 tDc2 i.xal
19 :xal f5 20 tDg5 'ii'a4 is analysis by
Schneider, who assesses this as a little
better for Black. A nice solution.
c) 1 1 a4 and now:
c 1 ) l l . . .ne8 1 2 h3 i.xf3 1 3 i.xf3
c4 ! ? 1 4 i.e3 �a5 1 5 .id4 J:te7 ! 1 6
�h l ( 1 6 'it>h2 tDc5 ! ? 1 7 e5 tDe8 =)
16 ... a6 17 g4 l:tae8 1 8 g5 tDxe4 ! with
good compensation for the exchange,
Peev-Velimirovic, Sofia 1 972.
c2) 1 l ...l:tc8 ! (more reliable) 12 h3
i.xf3 13 i.xf3 c4 14 .ie3 tDc5 15 e5
dxe5 16 fxe5 tDfd7 17 e6 tDe5 ! is
given by both Nunn and Litmanovic.
Kapengut then continues 1 8 i.xc5
.uxc5 19 tDe4 tDxf3+ 20 �xf3 fxe6 2 1
�g4 Ihfl+ 22 l:txfl J:.xd5 23 �xe6+
�h8 24 :f7 WVb6+ +.
d) 1 1 h3 ("the most frequently
played continuation", according to
Kapengut) 1 1 . . . .ixf3 1 2 i.xf3 l:tb8
(this move goes unmentioned by Kap­
engut, but I like it here, still waiting to
77
decide upon ...:cte8 or ...tDe8; 1 2 ...tDe8
is also OK, often transposing to note
'c' to White's 1 2th move) 1 3 :e l tDe8
(as in the main line, Black prepares
. . .tDc7 and . . .b5) and now:
d l ) 14 i.g4 f5 15 exf5 gxf5 1 6
i.e2 a6 1 7 i.d3 ( 1 7 a4 tDc7 1 8 a5
i.d4+ and ...�f6 - Va"isser) 17 ...tDc7
1 8 WVc2 �f6 19 tDe2 (versus . . . c4),
Monin-Schekachev, St Petersburg 1 994.
Here 1 9 . . Jlbe8 appears natural, but
1 9 ...tDxd5 is more ambitious. Then af­
ter 20 tDg3 Va"isser gives 20 . . .tDc7, but
it's not clear how he would answer 2 1
i.xf5 . On the other hand, 20. . .'iVh4 !
pins the knight and puts pressure on
the kingside; for example, 2 1 'it>h2
tDb4 (2 l . ..tD7f6 22 i.xf5 'it>h8 ! ?) 22
�b3+ 'it>h8 23 .id2 tDxd3 24 �xd3
tDf6 ! hoping for 25 tDxf5? (25 'ii'xf5
h5 ! ; 25 i.c3 ! d5 ! and Black seems
slightly better) 25 ... tDg4+ 26 'it>gl
.l::txf5 ! , etc.
d2) 14 a4 a6 (Va"isser's 14 ...tDc7 is
more flexible) 1 5 a5 tDc7 1 6 'iVd3 .l:.e8
1 7 .ie3 b5 1 8 axb6 :'xb6 1 9 :te2
�b8 ! 20 l:.a2 :b3 = Cebalo-Kristic,
Pula 1 999.
We now return to 1 1 Itel (D):
11 tDe8
.•.
This rather irregular idea ( 1 1 .. .:e8
is more common) has done well in
practice. Black's idea is to cover e5
with his g7 -bishop while preparing
queenside expansion via ... a6, ... tDc7
and ... b5.
Nevertheless, it would be nice to
temporize, waiting for h3, when White
loses a tempo. One attempt to do this,
l l . ..a6?, fails to 12 e5 ! tDe8 1 3 e6 fxe6
14 tDg5.
78 THE GAMBIT GUIDE TO THE MODERN BENONI
But I wonder about 1 1 ...�b8 ! ?, to
discourage eS and prepare ...bS. Then,
for example, 1 2 a4 ( 1 2 h3 i.xf3 1 3
i.xf3 ttJe8) 1 2. . .a6 1 3 as ( 1 3 h3 i.xf3
14 i.xf3 ttJe8) 1 3 ... ttJe8 could follow,
with a good version of the text, or per­
haps 1 3 ...�c7 and ...I:tfb8, now that eS
has lost its force.
This is just analysis, but I think that
1 1 .. .�b8 may be a very efficient solu­
tion to Black's problems. Compare
what follows.
12 ttJg5 (D)
Probably the most promising move.
Here are the alternatives:
a) 12 ttJd2 i.xe2 1 3 �xe2 a6 14 a4
i.d4+ I S 'iio>h 1 ttJef6 1 6 �f3 ( 1 6 ttJc4
ttJb6 =) 1 6...,Ue8 1 7 ttJe2 �e7 1 8 ttJxd4
cxd4 1 9 b3 ttJxdS 20 i.a3? (20 i.b2
ttJb4 ! +) 20...ttJe3 2 1 :ec 1 and now
21 ...ttJcS ! is even better than 2 1 ...�e6,
as played in Banikas-Beliavsky, Tyn­
iste 1 995 .
b) 1 2 i.e3 a6 1 3 a4 (or 1 2 a4 a6 1 3
i.e3) 1 3 . . J:tb8 ( 1 3 ...ttJc7 1 4 iLf2!? .i;lb8
I S i.h4 i.f6 1 6 i.g3 i.xf3 1 7 i.xf3,
Vatsser-Smirin, Moscow rpd 1 996,
and now 17 . .. i.d4+ ! 18 'iio>h l I:te8 with
counterplay - Vatsser) 14 as ttJc7 I S
ttJd2 i.xe2 1 6 �xe2 bS 1 7 axb6 ttJxb6
1 8 'i!i'd3, Flear-Mortensen, Reykjavik
ECC 1 999, 1 8 ... fS ! ? ( 1 8 ...:e8 =) 1 9
eS :e8 20 ttJc4 ttJxc4 2 1 �xc4 l:txb2
22 exd6 ttJbS with a slight advantage
for Black - Mortensen.
c) 1 2 h3 (the most common move)
1 2... i.xf3 1 3 i.xf3 ttJc7 ( 1 3 ... 1:tb8 is
note 'd' to White's 1 0th move) 1 4 a4
( 1 4 iLe3 l:tb8 I S i.g4 fS 1 6 exfS gxfS
1 7 i.f3 b5 ! 1 8 �c2 ttJb6! =) 1 4 ... a6 1 S
g4 :b8 1 6 gS (this expansion i s typi­
cally rather slow) 16 ... bS (or 1 6... fS)
17 axbS axbS 1 8 i.g4 b4 1 9 ttJe2 ttJbS
20 'i!i'c2 .l:te8 = Kniest-Glek, Berlin
1997.
B
12 i.xe2 13 .l:txe2!
.•.
Considered best, thinking about
'iWe l , and perhaps i.e3-f2. 1 3 'iWxe2 is
at best harmless; e.g., 1 3 ... a6 14 a4
i.d4+ ! I S 'iio>h 1 ttJef6 + Elbilia-Renet,
Manila OL 1 992, in view of 16 ttJf3
.i;le8 17 �d3 ttJg4 1 8 ttJd 1 l:.c8 ! 1 9
ttJxd4 cxd4 (Hebert) and Black has all
the squares.
13,..a6
 PAWN-STORM SYSTEMS
The immediate 1 3 . . .liJc7 is also
playable. Then, in line with the plan in
the last note, 14 'ii'e U te8 1 5 .i.e3? ( 1 5
.i.d2 =) allows 15 ...h6 16 liJf3 .i.xc3
17 bxc3 ':xe4, when White lacks com­
pensation.
14 a4 liJc7 15 'iid3
To stop ... b5. An alternative is 1 5
�el lte8 1 6 .i.d2 b5 1 7 axb5 axb5
with equality.
15 liJf6!?
.•.
1 5 ...ltb8 1 6 a5 b5 17 axb6 ':xb6
"also looks OK for Black" - Morten­
sen. Hebert even likes 1 5 ...b5 ! 1 6 axb5
axb5 1 7 ltxa8 'iixa8 with the idea that
1 8 liJxb5 is met by 1 8 ...'iia6.
16 b3 b5 17 .i.b2 bxa4
Often the best way to open lines
against the b3 and a4 structure.
18 liJxa4 ltb8 =
Chabanon-Degraeve, French Ch
(Meribel) 1 998 continued 1 9 liJf3 l:te8
20 g3 'iVd7 ! 2 1 ltd l ? ! (2 1 liJd2)
21 ...'ii'b5 22 .i.xf6 .i.xf6 23 e5 liJxd5
(23 . . .'ii'xb3 I ?) 24 'iixd5 'ii'xe2 25 exf6
�e3+ 26 <3;g2 ':e6 27 lld2 ':xf6 +.
White's pieces are coordinating poorly.
Black's l 1 . ..liJe8 idea has its pluses
and minuses, but it should hold the
balance and it avoids the complex the­
ory associated with the main lines of
9....i.g4. One might want to look into
1 1 ...Vib8 as well, since it defers the
... liJe8 decision while forcing White to
show his hand.
A22)
9 .liJbd7 (D)
..
A move enthusiastically endorsed
by Schneider, who gives it ' ! ' . I like it
as well.
79
W
10 e5
The only move that Schneider con­
siders, and certainly the most exciting,
but it may be that the direct 1 0 0-0 is
actually better. Then after 1 0 ... ':e8,
White has a number of ways to react to
the threat on e4:
a) 1 1 .i.d3 doesn't make much
sense: l 1 . . .c4 ! ? ( 1 1 . . .b5 ! ; l l ...a6 1 2
a4 Vic7 =) 1 2 .i.c2 b5 1 3 ':el a6 with
an equal position, Sznajder-Widera,
Pyskovice 1 99 1 .
b) 1 1 e5? ! dxe5 1 2 fxe5 liJxe5 1 3
liJxe5 .l:.xe5 1 4 .i.f4 .l:.e8 (or 1 4 ....:f5 !
1 5 g4 ':xd5 16 liJxd5 liJxd5 with an
obvious advantage for Black, Balogh­
Szalanczy, Budapest 1 99 1 ) 1 5 .i.f3
.i.f5 1 6 d6 'ili'd7 + Ishbulatov-Toth,
Budapest 1 998.
c) 1 1 'ii'c2 is recommended by
Va"isser, but I think that Black is hold­
ing his own after 1 l ...'iIi'e7 ! ( 1 1 . ..a6 1 2
a4 'ili'e7 ! ?) 1 2 ':el ! ( 1 2 liJd2 liJb6 ! ; for
example, 1 3 a4 liJfxd5 ! 1 4 exd5 .i.f5 !
15 'ili'b3 .i.xc3 and . . .'ii'xe2; 1 2 liJb5
liJxd5 ! 1 3 exd5 'ii'xe2 +) 1 2 ... liJb6 ! 1 3
a4 ( 1 3 h 3 .i.d7) 1 3 .. . .i.g4 = , with the
idea 14 a5 .i.xf3 1 5 gxf3 liJbd7.
80 THE GAMBIT GUIDE
d) I I lbd2 a6 1 2 a4 ( 1 2 lbc4 lbxe4
1 3 lbxe4 Itxe4 1 4 lbxd6 l:f.d4) trans­
poses into a variation that is well
known from the 9 ... l:f.e8 10 lbd2 move­
order. Now 1 2 ...c4 has been analysed
to death, but I like the less forcing op­
tion 1 2 .. Jlb8. Then:
d l ) 1 3 'ii'c2 lbb6 ( 1 3 ... hS is also
good) 14 h3? ( 14 as lba8 is equal)
14 . . .lbfxdS I S exdS i.. x c3 1 6 bxc3
I:l.xe2 += Dura-Martie, COIT. 1 987.
d2) 13 'i!fb3 is met by the familiar
1 3 . .. bS ! 14 axbS axbS I S lbxbS ( 1 S
�xbS lbg4 !) l S ...lbxe4 1 6 lbxe4 l:f.xe4
+ Schneider.
d3) 1 3 'it'h 1 (this move is the rea­
son that 1 2 ...,Ub8 has met with disap­
proval in the books) 1 3 ... lbb6! (a nice
improvement on the previous unsuc­
cessful moves 1 3 ...b6, 1 3 .. :iVc7, and
1 3 ... c4) 14 .l:.e1 ( 1 4 aS lba8, as always,
has the twin ideas of ... bS and ...lbc7-
bS) 14 ... i.d7 ( 1 4 ...hS ! ?) I S aS lba8 1 6
i.f3 hS ! (with the idea . . .lbg4) 1 7 h3?!
lbg4 ! 18 �e2 'iVh4 1 9 l:f.f1 i.d4 20
.ixg4 hxg4 2 1 �e1 'i!fxe1 (2l. . .�hS !
is very strong) 22 Itxe1 bS 23 axb6
4Jxb6 24 .uxa6 'it'g7 ! ? 2S 4Jb3 i.c8 26
I:ta7 i.xc3 27 bxc3 lbxdS 28 lbaS
lbxc3 ! 29 lbc6 lbxe4 -+ Apresa-Grau
Ribas, COIT. 1 994-2000.
We now return to 10 eS (D):
lO dxe5
...
Although not leading to such flashy
play, 1O ... lbe8 ! is a very attractive
move, and depending upon the assess­
ment of 13 .. .fxe6 in the main line be­
low, it may be objectively Black's best
continuation. White has to defend his
overextended pawns, and may even
have trouble fully equalizing:
TO THE MODERN BENONI
B
a) 1 1 e6? ! (given an ' ! ' by Schnei­
der, and the only move which has been
seen here, but it may well be a mis­
take) l l . ..fxe6 1 2 dxe6 ( 1 2 lbgS lbb6
1 3 lbxe6 i.xe6 1 4 dxe6, Kilpi-Lahti,
Jyvaskyla 1 997, and now 14 ...lbc7 I S
.i.g4 l:f.e8 i s probably the easiest path
to an advantage) 1 2...lbb6 1 3 a4 i.xe6
14 as lbc8 I S lbgS lbc7 1 6 lbxe6
lbxe6 1 7 'i!fdS l:i.e8 ! ? ( l 7 ...�h4+ ! 1 8
g3 �e7 1 9 i.c4 �h8 ! is a very strong
alternative) 1 8 0-0 'it'h8 + Kluss­
K.Kaufmann, Lublin 1 993. Schneider
suggests 1 9 'ii'xb7 lbe7 20 i.c4, but
20 ...dS ! refutes this, as both 2 1 lbxdS
'ub8 and 2 1 i.xdS l:tb8 22 �xb8 'ii'xb8
23 i.xe6 lbfS are pretty awful for
White.
b) 1 1 exd6 is quite easy to meet:
1 l . ..lbxd6 1 2 0-0 l:f.e8 + with moves
like ... lbb6 and ...i.fS or ...lbfS in the
air.
c) 1 1 .i.e3 ! ? looks better: 1 l ...�e7
12 �d2 ! dxeS 1 3 fxeS lbxeS 14 0-0-0! ?
intending to meet 1 4. . .lbg4 with I S
i.gS. This is very messy; my feeling is
that White has slightly less than full
compensation for the pawn.
 PAWN-STORM SYSTEMS
11 fxeS ctJg4 12 e6!
Vai'sser suggests 1 2 �g5, which I
think should be answered by 1 2...'iib6
1 3 ctJa4 ( 1 3 e6 'iixb2 1 4 0-0 'iixc3 1 5
exd7 �xd7 -+; 1 3 0-0 ctJdxe5 i s better
for Black than the similar 9 . . .l:.e8 lines,
because f7 is covered) 1 3 .. :�i'a5+ 14
i.d2 'fIc7 + (the centre is falling).
12 ctJdeS!
...
Ambitious, and Schneider's main
move, even though his continuation is
rather dubious. The alternative 12.. .fxe6
1 3 dxe6 ctJde5 1 4 'iixd8 ':xd8 is pas­
sive, but not so bad. Accurate play
yields White a very small edge, but all
the games from this position have
been drawn thus far. Of course, Black
would like to play for more.
13 ctJgS (D)
Certainly the critical continuation,
as shown by the following lines:
a) 1 3 exf7+ l:.xf7 14 O-O? ( l 4 ctJg5
l:tf5 1 5 ctJce4 is assessed as 'unclear'
by Stohl; in that case, 1 5 . . .h6 1 6 ctJe6
.i xe6 1 7 dxe6 'ii'xd l + is at least equal)
1 4 . . . ctJxf3+ 1 5 �xf3 �d4+ 1 6 �h l
etJxh2 1 7 ctJe4 ctJxf3 1 8 :'xf3 'ii'h4+
0- 1 Kalousek-A.Toth, Budapest 1 998.
b) 13 �f4 ? fxe6 1 4 ctJxe5 ctJxe5 1 5
0-0 exd5 1 6 'ii'xd5+ 'ii'xd5 1 7 ctJxd5
.ig4! + Kahn-A.Schneider, Budapest
1 992.
c) 1 3 ctJxe5 ctJxe5 14 exf7 + lhf7
1 5 .te3 ctJc4! (Schneider's suggestion,
to improve upon his own 1 5 . . :ij'h4+
1 6 g3 'ii'e7 17 �d2 + Szabolcsi­
A. Schneider, Budapest 1 993) 16 �xc5
( 1 6 i.xc4 "iYh4+ +) 1 6 . . . ctJxb2 1 7
'i'b3 i.xc3+ 1 8 1i'xc3 'ii'xd5 1 9 �d4
.ig4 ! ! 20 �xg4 l:.e8+ 2 1 �d2 'iixg2+
-+ (all analysis by Schneider).
81
B
13 c4!!?
...
A beautiful and compelling move
originally played by I.Zaitsev. Schnei­
der assigns it a ' ! ! ' , not unjustly, but in
the end I question its ultimate value. If
a pessimistic assessment of 1 3 . . . c4 is
right, the best move here may actually
be the promising sacrifice 1 3 . . . fxe6 ! ?
1 4 �xg4 ctJxg4 1 5 "iYxg4 exd5 with
two pawns for a piece. This has only
been played a few times, and is hard to
assess on that basis alone. However, I
tend to favour Black because White's
king ends up in the centre fighting the
bishop-pair and he hasn't any obvious
counterplay. 16 'iYh4 ( 1 6 "iYg3 l:te8+
17 �d l �f5 intending . . ."iYd7 and a
devastating advance of pawns in the
centre) 16 . . .h6 17 ctJf3 g5 and now:
a) 1 8 �xg5 hxg5 1 9 "iYxg5 'iYxg5
20 ctJxg5 �g4 is fine for Black.
b) 1 8 "iYg3 d4 1 9 ctJbl ? ! 'iVe8+ 20
�d l ?, as in Pliasunov-Khismatullin,
Russian U- 1 6 Ch (St Petersburg) 1998,
loses to 20 ... d3 ! 2 1 l:te l 'iYa4+.
c) 18 "iYh5 !? is untested, but equally
depressing for White after something
like 18 ... d4 1 9 etJe2 "iYe7.
82 THE GAMBIT GUIDE TO THE MODERN BENONI
d) 1 8 'iia4 g4! ? 1 9 liJd2 i.d7 20
1Vb3 c4! 2 1 'iWxb7 'iie7+ 22 liJe2 l:be8
23 'iixdS+ 'ifilh8 24 'iixc4 ':'f4 ! and
Black is winning, Kopionkin-Ulko,
Smolensk 2000.
I suspect that 1 3 . . .fxe6 is objectively
good for Black. At any rate, White's
practical chances of defending suc­
cessfully are not good at all.
14 0-0!
Or:
a) 14 i.xg4 liJd3+ I S 'ifile2 fxe6 1 6
i.f3 exdS 1 7 i.xdS+ 'ifilh8 1 8 l:.f1
i.g4+ 19 i.f3 i.xc3! 20 i.xg4 'iid4 !
with a winning attack (I.Zaitsev).
b) 1 4 liJxf7 liJxf7 ! ( 1 4 ... 'iib6 I S
liJe4 liJxf7 and now 1 6 exf7+ ? loses to
1 6 ... ':'xf7 1 7 i.xg4 i.xg4 1 8 'iixg4
l:.e8, whereas 1 6 i.xg4 liJd6! is better
for Black, but not clearly so) I S i.xg4
( I S exf7+ ':'xf7 1 6 i.xg4? 'iih4+ 1 7
g3 1Vxg4 + ) I S ... .l.xc3+! 1 6 bxc3 liJeS
17 i.e2 (there is no good move; for ex­
ample, 1 7 .l.f3 liJxf3+ 1 8 gxf3 ':'fS ! )
1 7 . . . 'iih4+ 1 8 g3 'iWe4 1 9 l:.f1 i.xe6!
with a decisive advantage for Black.
14 1Vb6+ 15 'ifilhl liJf2+ 16 l:.xf2
.•.
'iixf2 17 liJge4 'iib6 18 e7 l:.e8 19 d6
(D)
19 liJd3!?
.•.
This is Schneider's attempted im­
provement upon 1 9 . . . 'iic6 20 liJdS
'ith8, Knezevic-I.Zaitsev, Smederev­
ska Palanka 1 97 1 , when 2 1 liJc7 !
should be very good for White, al­
though this is not initially obvious.
In my opinion, Black's best chance
to save Zaitsev's 1 3 ... c4 is probably
1 9 . . . i.e6. Then 20 liJdS i.xdS 2 1
'iixdS i s still extremely difficult to an­
swer; the best try for Black may be
B
2 l .. .'iib4 22 i.d2 'ii'xb2, although I
still can't find a good solution to 23
l:.f1 with multiple ideas such as liJgS,
i.g4, i.c3 and d7.
20 i.xd3 cxd3 21 1Vxd3 i.f5 22
liJd5! 'iix d6!?
Thus far Schneider's analysis. He
continues 23 liJdf6+ 'iixf6 24 liJxf6+
.l.xf6 2S 'iib3 ':'xe7 26 .l.d2 ':'ae8 +,
which is fair, but sadly, 23 liJxd6 i.xd3
24 i.f4 ! intending liJc7 is clearly better
for White, since Black cannot deal with
the e-pawn.
The fate of the melee introduced by
l l . ..liJg4 appears to rest mainly with
1 3 ... fxe6. However that turns out, it
seems to me that Black's option of
1 0. . . liJe8 means that he stands well
enough after 10 eS . Since Black's play
after 1 0 0-0 also looks satisfactory, it's
strange that 9 . . .liJbd7 has been ne­
glected for so long.
A3)
8 i.b5+ (D)
This is the Taimanov Attack, a
greatly feared weapon that (along with
some other lines, such as the i.d3 and
 PAWN-STORM SYSTEMS
ttJge2 1ine of Chapter 7) has driven nu­
merous Benoni players away from the
'pure' 2 . . . c5 3 d5 e6 4 ltJc3 move­
order. Many, if not most, Benoni ad­
vocates now wait for White to commit
to ltJf3, as in the sequence 1 d4 ltJf6 2
c4 e6 3 ltJf3 c5 4 d5 exd5 5 cxd5, etc.
If White plays 3 ttJc3 with this move­
order, Black can choose to play the
Nirnzo-Indian (3 ...i.b4) or the Queen' s
Gambit Declined (3 ... d5). These days,
a lot of people don't even like the
Benoni with ltJf3 in, due to the Mod­
em Main Line (Chapter 9). However, I
think that I have provided good ways
to meet that system, and I will try to do
the same here.
This is a particularly difficult task,
given the space required to examine
two systems in detail. Therefore, I
have split up the work into two parts.
One is the presentation in this section,
which itself is a double repertoire due
to the option of . . .'iVd8 or . . .'iVe7 ideas.
The other is an 'avoidance' move­
order, trying to bypass the problem
before it arises. The latter is risky and
completely experimental, in contrast
83
to the main repertoire lines, which fol­
low immediately below.
I wrote a book about the Taimanov
Attack in 1 985 that reflected my great
optimism about it. I have also played it
myself, with good results. Established
theory remains favourable to White,
although ECO rather surprisingly gives
equality in one commonly played line
involving ... ltJa6-c7. I ' m a little
sceptical of that and in fact, my oId
book is still relevant in calling that line
into question. What I propose here is
to play a different set of lines involv­
ing .. .'iii'h4+. This idea achieved some
popularity after I wrote my book, and I
feel that it is underrated. Black takes a
practical approach to the position, and
the strategies tend to be well-defined.
At worst, the . . .'ilVh4+ lines should
limit White's advantage to about what
he achieves in most openings. I will
try to show that they do more than that.
Before continuing, I should note
two things that apply to this section.
Perhaps more than any other variation,
the Taimanov Attack will test Black's
defensive and counterattacking skills.
These positions are more resilient than
they might at first appear, and one has
to keep in mind the hypermodern prin­
ciple that White's advanced central
pawns are not only menacing but also
weaknesses. Secondly, many of the
positions that arise in this section are
unbelievably complicated. One can
pursue a complex line in detail for ten
moves only to reach a position that is
quite as chaotic as the one you started
with. I have done a mass of speculative
analysis here, to the best of my ability,
84 THE GAMBIT GUIDE TO THE MODERN BENONI
but the reader should be aware that
some of it will unquestionably be sub-
ject to improvement. My advice would
be to study and reflect upon the mate-
rial rather than to try to commit a great
deal of it to memory.
8. ttJfd7
..
In the end, despite enormous com­
plications, I think that 8 ...ttJbd7 is just
unsound, and 8 ... i..d7 yields White a
clear advantage. After the correct move
8 ... ttJfd7, let's see how White can play
it:
A31: 9 ttJf3 84
A32: 9 i..e2 86
A33: 9 i..d3 87
A34: 9 a4 92
The last move is the one that has
discouraged so many Benoni players.
A3 1 )
9 ttJf3
This move has become more popu­
lar lately, but 1 don't see it as too chal­
lenging, since Black can now expand
on the queenside by ... a6 and ...b5.
9 a6 10 i..d3
•.•
In this position, 1 0 i.. xd7+? is quite
illogical, but has been played more of­
ten than one would ever imagine, with
an overwhelming score for Black after
1O ... ttJxd7 (see the note to White's 8th
move in Line D2 of Chapter 1 ). And
10 i..e2 b5 doesn' t make much sense
either, as White's e-pawn lacks protec­
tion. Polikarpov-Kapengut, Minsk
1976 continued 1 1 a4 b4 12 ttJbl 0-0
1 3 0-0 ttJf6 14 ttJbd2 1:!e8 15 i..d3 i..g4
16 h3 i..xf3 17 'iVxf3 ( 1 7 .l:i.xf3 c4 !)
17 ... ttJxd5 ! 18 exd5 .td4+ and . . .:te3.
10 b5 11 0-0 0-0 (D)
•.•
W
12 'it>hl
This is played in about 70% of the
games with 9 ttJf3 . White wants to
see what Black is doing, and dodges
checks along the diagonal ( ...c4 and
. . .'ii'b 6). He should get nothing special
from 12 i..e3 ttJf6 or 1 2 a3 ttJb6. The
following alternatives are a bit more
challenging, but probably of equiva­
lent worth:
a) 1 2 'iie l l:te8 1 3 �g3 c4 1 4 i..c2
b4 1 5 ttJa4 ( 1 5 ttJd 1 ttJc5 16 ttJf2
ttJbd7 17 ttJg5 ! ? ttJf6 1 8 i..e 3? ttJh5 1 9
"iYf3 h 6 + Suveges-Retter, corr. 1 996)
15 . . .ttJf6 16 f5 i..d7 ( 1 6 . . .ttJxe4 1 7
i..xe4 :xe4 may well be good) 17 i..g5
i..b5 = Cherepkov-Suetin, Sochi 1 96 1 .
b) 1 2 .:te l b4 ! ? ( I like 1 2. . .ttJb6!
with the idea . . .i..g4 and/or ... c4 and
. . .b4) 1 3 ttJa4 ttJf6? ( 1 3 . . J�e8 !) 14 h3?
(White should play 14 e5 ! with the
idea 1 4 . . .dxe5 15 fxe5 ttJxd5 1 6 .i.g5
ttJe7 17 ttJxc5 ±) 14 ...i..d7 15 a3 l:te8
16 axb4 cxb4 17 l:ta2 i..b5 18 b3 l:ta7
= intending .. J�ae7, Nechaev-M.Kopy­
lov, Donetsk Z 1 998.
c) 1 2 f5 c4 ! ? (giving up d4 like this
can be risky; 12 . . .ttJe5 =) 1 3 i..c2
 PAWN-STORM SYSTEMS
tbcS? ! ( I 3 ...tbeS ! 14 tbxeS .i.xeS is
still equal) 14 .i.gS ! 'iib6, T.Reich­
Bauml, Bad Worishofen 2000, and
now I S .i.e3, intending tbd4 or .i.d4 at
some point, is better for White.
12 ...tbb6
Hebert highlights this move in his
Benoni CD, and I think that it's an ap­
pealing one, since it covers c4, makes
room for . . .tb8d7, allows . . . .i.g4 in
some lines, and even opens up the idea
of .. .l::ta7 followed by .. .':c7 or ...l:te7.
12 . .J::t e8 is the most common move
and it is perfectly playable, but Black
has to be careful after fS, since the
rook is no longer defending f7. Black
has also played 1 2...tbf6, 1 2... c4, and
even the immediate 1 2 ....l:.a7.
13 f5
The only move I've found in data­
bases - it's very logical to prevent
... .i.g4 and free the c l -bishop at the
same time.
13 b4
.•.
Driving the knight towards the king­
side may not be best. A good line is the
straightforward 1 3 . . . tb8d7 14 .i.gS
( 1 4 fxg6 fxg6 ! is an original approach:
IS tbgS liJeS 1 6 .l:.xf8+ 'iVxf8 17 .i.e2
WIIe7 + Dargena-Delarge, COIT. 1 985)
14 ... .i.f6 IS .i.f4 'iVe7 ( I S ... b4 ! ? 1 6
tba4! liJxa4 1 7 'iVxa4 .i.xb2 1 8 l:tabl
i..eS 19 liJxeS dxeS 20 .i.h6 with un­
clear compensation) 16 'iVd2 .i.b7 1 7
.:i.ael .l:.ac8 = Glek-Anikaev, Minsk
1983.
14 tbe2 liJ8d7
An incredibly messy game that I
am afraid to comment upon continued
14 ... aS I S .i.gS .i.f6 1 6 'iVd2 lla7 1 7
WIIf4 .i.xgS 1 8 liJxgS h 6 1 9 liJh3 g S 20
85
�g3 'iVf6 2 1 liJf2 l:Id8 22 liJg4 'ii'h8 23
eS liJ8d7 24 f6 liJxeS 2S liJxeS dxeS
26 'iVxeS liJxdS 27 l:tadl .i.e6 28 .i.fS
.i.xfS 29 'iVxfS l:tad7 30 liJg3 1Wh7 and,
in Glek-Tataev, Moscow 1 992, Black
was doing well, although he later lost.
15 .i.g5 .i.f6!
This tends to be the best answer to
.i.gS in cases where Black has already
made progress on the queenside.
16 .i.f4
1 6 .i.h6 .l:.e8 is equal.
16...'iVe7 17 �d2 tbe5 (D)
Hebert mentions 17 ...c4 ! ? 1 8 .i.b 1
gxfS ! ? 1 9 'iVxb4 liJxdS 20 exdS �xe2,
but I don't like 2 1 .i.xfS for Black, or
even his continuation 2 1 liJd2 c 3 ! ? 22
bxc3 .l:.b8, due to 23 �xd6 .i.xc3 24
.i.xfS ! .
18 liJxe5 .i.xe5 19 .i.xe5 'iVxe5 20
f6!?
This is Schipkov-Szalanczy, Bu­
charest 1 993. Now Hebert suggests ei­
ther 20 ... gS ! ? "or the simple and
possibly even safer 20 ... 'it>h8 2 1 �h6
.l:.g8". Since Black has a long-term
positional advantage in the latter case,
86 THE GAMBIT GUIDE TO THE MODERN BENONI
White must move quickly, and Hebert
gives 22 ttJg l (intending ttJf3-gS; 22
':f4?? gS) 22 .. J!a7 ! , when instead of
his 23 ttJf3? ! 'iWxf6, White might try 23
l:tf4 (23 :f3 c4 24 .:tc l i.g4 !) 23 ... gS
24 l:tf2. Then a plausible sequence
might be 24 ...ttJd7 2S ':afl l:tg6 26
"iib3 g4 27 'iWh4 �gS 28 �xgS ':xgS
followed by ...ttJeS. This probably isn't
much, but I'd rather be Black.
In general, I don't think that allow­
ing ... bS can be White's best course.
A32)
9 i.e2 (D)
B 1 993.
Trying to get back to a Four Pawns
Attack with the knight misplaced on
d7. However, the bishop isn't well
placed on e2 except in positions where
eS is possible.
9 :ti'h4+
•.
Consistent with the themes of this
section. Black wants to weaken White's
kingside before continuing with his
development.
Naturally, 9 . . .0-0 1 0 ttJf3 lIe8 1 1
0-0 can't be bad:
a) A rare and complex idea is
1 l ...i.xc3 ! ? 12 bxc3 .l:.xe4, which most
players would shy away from (if only
for practical reasons) because White's
dark-squared bishop is unopposed.
However, this sequence might also
discourage some players of White, as
Black has won a central pawn. A game
which favoured Black went 1 3 i.d3
( 1 3 ttJgS lIe8 14 fS Hebert; then
14 .. .':tjf6 ! ? I S fxg6 hxg6 is complex)
1 3 .. J�e8 14 c4 ttJf6 I S i.b2 i.g4! 1 6
'ii'c2 ttJbd7 1 7 ttJgS 'ii'e7 ! (threatening
...i.e2) 1 8 �c3 �e3+ 19 l:.f2 ( 1 9 'iii>h l
i.e2 !) 1 9 ... i.fS ( 1 9. . .bS ! 20 h3 i.fS 2 1
i.xfS 'ii'xc3 22 i.xc3 gxfS 23 cxb5
ttJxdS +) 20 i.xfS "iVxc3 2 1 i.xc3
gxfS + Shafii-Sabitov, Kuala Lumpur
b) After l l .. .ttJa6, 12 'iii>h 1 has kept
the advantage in several games, but,
l.1vanov's proposal of 1 2...ttJb6! with
the idea of ... c4 and ...ttJcS highlights
the one-dimensionality of 9 i.e2. Then
1 3 i.e3? i.xc3 1 4 bxc3 .:txe4 com­
pares very poorly with line 'a' for
White, and 1 3 i.d2 c4 or 1 3 ttJd2 i.d7
poses Black no problems.
10 g3
We will see 10 'iii>f l and similar
ideas below. In this case, Black has the
interesting move 1 0. . .�e7 1 1 ttJf3,
when 1 l . ..i.xc3 ! ? 12 bxc3 'ii'xe4 ex­
ploits the fact that White can't mobi­
lize quickly with his king on fl . If that
looks too risky, l l . ..ttJb6 1 2 'iii>f2 i.g4
1 3 I1el ttJ8d7 is solid, intending 14 eS
O-O ! .
10 'iWe7
...
Or 1 0...'ii'd8 I 1 ttJf3 ( l l ttJbS ! ? ttJb6
12 eS dxeS 1 3 d6 ttJa6 14 fxeS, Paramos
 PAWN-STORM SYSTEMS
Dominguez-Reinaldo Castineira, Mon­
dariz 1 996, and now 1 4 . . . i.xe5 1 5
ttJf3 i.g7 would make White justify
his play) 1 1 . . .0-0 1 2 0-0 .l:te8 1 3 .l:te 1
ttJa6 ( 1 3 . . .i.xc3 ! ?) 14 i.f1 ttJb6 15 h3
c4 16 e5 ttJb4 ! 17 g4 (an odd move,
but the forcing 1 7 a3 allows 17 . . .ttJd3
18 .i.xd3 cxd3 1 9 g4 dxe5 20 'ili'xd3
e4 ! 2 1 .l:txe4 .l:txe4 22 'ili'xe4 f5 !)
17 ... dxe5 18 fxe5 ttJ6xd5 1 9 ttJxd5
'i'xdS 20 'ili'xdS ttJxdS 21 i.xc4 ttJb6 22
i.b3 .i.e6 =+= (pawn-structure) Borik­
Hort, Bundesliga 1 982/3.
11 ttJf3 ttJb6!? 12 0-0 i.g4 13 eS
Else just . . .ttJ8d7 follows, with
. . . .i.xf3 at the appropriate moment.
13 0-0 14 ttJe4
•.•
More accurate looks 1 4 ttJb5 ttJc8
1 5 i.e3 b6 =.
14...dxeS
14 . . .ttJc8 ! ? is still possible, to avoid
complications.
15 d6 'it'eS 16 fxeS ttJSd7
Korzubov-Kapengut, Minsk 1 985.
This has been assessed as equal, but it
may well be better for Black, since
White's centre is crumbling.
A33)
9 i.d3 (D)
This simple retreat was White's
main attacking try for many years, and
is still extremely important. Most re­
cent sources (e.g., Schneider, ECO
and NCO) give 9 i.d3 as ending in
equality or an unclear position. How­
ever, there are a variety of recommen­
dations as to how to do this, and no
definitive answer. For that reason, I
will examine 9 i.d3 in detail.
9.. �h4+!?
. considering for Black. Nevertheless,
87
Again this disruptive move is used
to weaken White's kingside at the cost
of a tempo. I didn't approve of 9 ...'it'h4+
when I wrote my book some years
ago, but now I think that the idea of
playing a later . . .i.g4 is much more ef­
fective than I realized. Black loses less
time than it seems after g3 and his
retreat to e7 or d8, since White has
weakened his kingside and has to deal
with a potential ... i.h3 or . . . i.g4. This
in tum takes time to counter, e.g., by
'it.i>g2 and h3.
Still, since 9 ... 'it'h4+ is a bit eccen­
tric, I should mention that theory con­
siders the 'main line' with . . . ttJa6-c7
equal. I'm not sure, but very briefly,
the primary line goes 9 . . .0-0 1 0 ttJf3
ttJa6 1 1 0-0 .l:tb8 1 2 'it.i>h l ( 1 2 'ili'el b5 !
1 3 ttJxb5 c4 14 i.xc4 1hb5 1 5 i.xb5
'it'b6+ =; 1 2 i.e3 ttJc7 1 3 a4 a6 14
i.f2 ! b5 1 5 i.h4 i.f6 1 6 i.xf6 ttJxf6)
1 2 . . .ttJc7 1 3 a4 a6 1 4 a5 b5 ( 1 4 . . ..l:te8 !
1 5 i.e3 b5 1 6 axb6 .l:txb6 =) 1 5 axb6
ttJxb6 1 6 f5 ( 1 6 i.e3 ! ? f5 1 7 e5 i.b7
=+=) 1 6. . .gxf5 17 exf5 ttJbxd5, which is
thought equal. This is certainly worth
88 THE GAMBIT GUIDE TO THE MODERN BENONI
White has other attacking ideas after
1 1 . . .l:tb8, and I'm not unreservedly
enthusiastic about it.
10 g3
1 0 �fl ! ? is extremely rare here.
Please refer to the �fl lines in Line
A34; for example, 1O ...Wid8 I l ltJf3 a6
12 a4 is analysed there.
10 Wie7
...
The most common move, but I think
that 1O ...'ii'd8( !) is definitely playable,
and probably even better. Although
Black loses a tempo (i.e., White's ltJf3;
as explained above, g3 hardly counts
as a 'tempo'), White's kingside is
weakened and Black's queen can go
to the queenside to apply pressure.
Also, in cases where ...ltJf6 or ...ltJb6
is played, White's eS is delayed since
his d-pawn is then exposed. Finally,
the ugly possibility of d6 with tempo
is eliminated. After 1 O .. .'iWd8, play
continues I 1 ltJf3 0-0 1 2 0-0 (D). One
might wonder, in comparison with a line
like 9 ...0-0 10 ltJf3 ltJf6 1 1 h3 (to stop
... .ig4), why White doesn't play 1 2 h3
here. But in the line before us, the h­
pawn is loose; e.g., 1 2 ...l:te8 1 3 0-0
(else ... fS) 1 3 ... .ixc3 ! 14 bxc3 ltJf6 =t.
From the diagram (see top of next
column), 1 2 ... a6 1 3 a4 transposes to
Line A34 1 3 (via the move-order 9 a4
'iVh4+ 1 0 g3 'ii'd8 1 1 ltJf3 0-0 1 2 0-0
a6 13 .id3). That is a recommended
repertoire line, but Black has some
nice alternatives with this particular
move-order:
a) 1 2 ...ltJb6 and now:
a l ) 1 3 a4 .ih3 (or 1 3 ....ig4 14 as
ltJc8) 14 l:tel .ig4 IS aS ltJ6d7 1 6 h3
.ixf3 17 'iWxf3 ltJa6 = .
B
a2) 1 3 �g2 (to stop ... .ih3 and
prepare h3) 1 3 ... i..g4 14 h3 .ixf3+ I S
'iVxf3 ltJ8d7 ( I will discuss this basic
type of position at length below, in the
note to Black's I Sth move) 1 6 lIel
l:te8. It's hard for White to make prog­
ress now; for example: 1 7 i.e3 lIc8 1 8
l:tac 1 a6 ! ? 1 9 .if2 ltJa8 20 a4 (20 eS
dxeS 2 1 fS e4! ? 22 i..xe4 ltJeS 23 'iff4
bS 24 l:tcd l ltJc4 2S 'ifc I ltJxb2 ! 26
'ifxb2 b4 with an excellent game)
20...c4 2 1 i..c2 'ifaS with an equal po­
sition.
b) 1 2 ...11e8 1 3 :tel ltJb6 14 'ifb3
( 14 �g2 is met by 14 . . .i..g4, as usual;
14 fS makes less sense with the rook
on e l instead of fl ; Black plays
14 ...ltJ8d7 =, which you should com­
pare with similar lines below) 14 ...i.g4
I S ltJd2 ltJa6 16 'Otg2 'iVc8 ! with plenty
of play; for example, 17 a4 ( 17 ltJbS
i..h 3+ 1 8 �h l 'iVd8 1 9 ltJf3 i..g 4! 20
�g2 ltJb4 with ...'ii'd 7, ... i.xf3+ and
...'ifh3 in mind, and answering 21 i..f l
b y 2 1 .. .ltJ4xdS) 1 7 . . . i..h 3+ 1 8 �h l
ltJb4 1 9 .ibS ltJd7 20 ltJf3 i.g4 2 1 l:tfl
a6 22 i..e2 fS ! .
We now return to 1 O. . .'iWe7 (D):
 PAWN-STORM SYSTEMS
l1 lbf3
1 1 'iie2 0-0 1 2 lbf3 tends to waste a
tempo for White in the . . .i.g4 lines
where Black plays . . . i.xf3. One game
went 1 2 . . .lbb6 1 3 0-0 i.g4 14 a4 as? !
(an odd move i n an odd game, paralys­
ing the queenside; Black has no prob­
lems after 1 4 ... lb8d7 or 1 4. . . a6 I S as
CDc8) I S i.d2 lb8d7 1 6 lbd l ? ! ( 1 6
\'i'g2 =) 1 6. . .I;'tfe8 1 7 'iig2 c4 1 8 i.c2
lbcs 19 .l:i.el 'iVd7 + hitting the a-pawn
and contemplating . . . fS, Hein-Bur­
meister, Germany tt 1 992/3.
11 ...0-0 12 0-0
Again, as in the line 1O ...'iVd8 1 1
CDf3 0-0 1 2 h3 (note to Black's 1 0th
move), the move 1 2 h3?! (to prevent
. . . �g4) is too slow. Among other
moves, 1 2 . . . fS ! at least equalizes.
12 lbb6
•.•
With the usual idea of . . . i.g4 or
. . . i. h3. It's important to notice that
1 2 . . . a6 1 3 a4 lbf6 transposes to Line
A3423, an alternative defence that the
reader might like. Here there are two
other moves which deserve attention:
a) 1 2 . . . lba6 1 3 .l:.e 1 ( 1 3 fS lbeS =
- compare note 'a' to White's 1 3th
89
move; 1 3 eS dxeS 1 4 fS is a standard
idea, but in this case 14 . . . lbb6 is a
good answer) 1 3 . . .lbb6 14 i.f1 ( 1 4 eS
i.g4 is equal) 14 . . . i.g4 1 S h3 i.xf3 1 6
�xf3 with a typical position that
should be equal. Black might even try
the odd 1 6 .. .fS ! ? ( 1 6 ... lbd7 would be
normal, followed by queenside expan­
sion) 1 7 eS dxeS 1 8 fxeS i.xeS 1 9
i.f4 i.d4+ 20 �h2 �g7 2 1 d6 l:tf7 ! ?,
when White has definite compensa­
tion, but probably no more than a
pawn's worth.
b) 1 2 ... lbf6 ! ? 1 3 �g2 ( 1 3 eS lbe8
14 'iib3 iLg4 is again very unclear; 1 3
.:tel i.g4 1 4 eS lbe8 I S e6 fxe6 1 6
dxe6 lbc7 1 7 i.c4 lbc6 +) 1 3 . . . i.g4 1 4
h 3 i.xf3+ I S 'iixf3 lbbd7 =; e.g., 1 6
i.d2 a6 1 7 a4 l:tfc8!? 1 8 b 3 l:tab8 1 9
l:tae 1 lbe8 ! = intending . . .lbc7 and
. . .bS . See the note to Black's I Sth
move for general comments about this
structure.
We now return to 1 2 . . .lbb6 (D):
13 �g2
This is the most frequent move in
such positions, preventing ...i.h3 and
90 THE GAMBIT GUIDE TO THE MODERN BENONI
preparing h3. However, the alterna­
tives deserve investigation:
a) 1 3 fS ! 7 is a dangerous attacking
move (which, however, gives up eS):
1 3 .. .lb8d7 14 ..tgS ..tf6 ( 1 4 ...f67 ! I S
..tf4 lbeS ;1; ; Black's miserable g7-
bishop guarantees White the advan­
tage) I S ..th6 :e8! ( 1 S .....tg7 1 6 iLxg7
<3;;xg7 17 'iWd2 with an attack) 16 <3;;h 1
lbeS ( 1 6 . . . gxfS 1 7 lbh4 !) 17 lbxeS
..txeS 1 8 'iWd2 and now:
a l ) After 1 8 . . . c47 (a premature
move, following Black's good play so
far), Ravikumar-Thorsteins, Copen­
hagen 1 982 continued instructively:
1 9 ..tc2 lbd7 ( 1 9 . . . ..td7 20 :If2) 20
l:tf2 ! a6 2 1 :afl bS. Now White could
have played 22 fxg6 ! fxg6 23 :f3, in­
tending 'iWf2, thereby gaining a large
advantage.
a2) Black can keep the balance
with the solid 1 8 . . . f6, when White
finds it difficult to continue the attack.
a3) Black can try the more ambi­
tious 1 8 ... ..td7 !7, preserving the pos­
sibility of ... ..td4 among other ideas.
Play might then continue 19 :f2 gxfS
and now 20 :Iafl f6 or 20 exfS ..txc3 ! 7
(20. . . f6 ! 7) 2 1 'iWxc3 f6 22 l:!.f4 <3;;h8 =.
Black is threatening . . .lbxdS (espe­
cially in conjunction with a potential
. . . ..tc6), and his good bishop on d7 can
support a queenside advance.
b) 1 3 a4 is Nunn's suggestion, to
which I gave an ' ! ' in my book. How­
ever, 1 3 . . ...tg4 1 4 as lbc8 intending
. . .lbd7 is a type of position we will see
in very similar examples below, and
not at all bad for Black.
c) 13 :tel ..tg4 1 4 ..tfl was played
in Sliwa-Gromek, corr. 1960. Then, as
well the natural 14 ...lb8d7, Black can
try 14 ... ..txc3 I S bxc3 ..txf3 16 �xf3
lbxdS =, when his extra pawn and
solid pawn-structure compensate for
the bishops. An example would be 1 7
..tb2 ( 1 7 ..td2 lbc7) 17 . . .lbc7 1 8 �ad l
( 1 8 c4 lbc6 1 9 'uadl lbe6) 1 8 ... lbc6.
13 ..t g4
•.•
1 3 . . .lba6 has been played here, but
I don't like 1 4 h3 ! c4 I S ..tc2 lbcs,
when 16 eS ! dxeS 17 fxeS is not attrac­
tive for Black, since 1 7 . . . ..txeS? is met
by 1 8 lbxeS 'iWxeS 1 9 ..tf4 'i!ie7 20
'i!id4 ! . Nevertheless, perhaps Black
can improve here.
14 h3 ..txf3+ 15 'iWxf3 lb8d7 (D)
Black can also play for immediate
activity by I S ... c4 16 ..tc2 lba6 1 7 a3
lbcs. This gives up d4 in return for
queenside play. Lau-Dolmatov, Graz
U-26 Wcht 1 9 8 1 continued 1 8 ..te3
lbbd7 1 9 :adl l:!.ab8 20 �fel bS 2 1
eS ! b4 (2 1 . ..dxeS 22 d6 'i!id8 23 'i!id5 ! )
2 2 axb4 1hb4 2 3 ..tc l ! '12_ 112. Kapen­
gut rightly calls this ;1;, but Black can
hold on after 23 . . .dxeS 24 d6 'i!id8 25
lbdS 1:.bS ! 26 lbe7+ <3;;h 8 27 lbc6
�a8 ! '
w
 PAWN-STORM SYSTEMS
Since this general type of position
arises quite often in these lines, I
should say some words about it. One
should be aware that, given a knight on
d7, Black's other knight may also be
on e8, c7, or even (after White's a4) on
b4, and the general discussion won't
change that much. One should also
note that, because of White's g3, he
had to waste an extra move (�g2) to
achieve h3 in comparison with similar
Pawn-Storm lines without ...�h4+.
Black's initial idea is prophylactic,
that is, to prevent White from making
pawn advances and freeing his two
bishops. Thus his exaggerated con­
centration of forces versus eS, for that
move would tend to unleash all of
White's pieces. If Black succeeds in
limiting White's ideas, he can then
turn his attention to advancing his own
natural majority on the queenside,
supported by rooks and a knight. This
advance, in conjunction with the pow­
erful bishop on g7, will ultimately
wreak havoc on White's queenside.
Crucially, that attack will also create
natural outposts for Black's knights,
which for the moment serve a defen­
sive role. As in many Benoni varia­
tions, it's difficult for White to keep
queenside lines closed indefinitely.
The familiar plan of :bl and b4 is not
very practical given the disposition of
both sides' forces.
So what can White do in such posi­
tions? Sometimes he can play eS any­
way, sacrificing a pawn for activity.
Assuming that Black has prevented
that, White could always play fS, but
this gives Black a grip on the dark
91
squares and in particular, the eS-square,
e.g. for . . . lbeS . Although fS must al­
ways be taken into consideration, it
tends to be a better move if it limits the
scope of a light-squared bishop on c8
or d7, which in this variation has al­
ready been exchanged. One might be
tempted to play g4 instead, but in
many cases that can weaken the king­
side. For one thing, White has to be
careful of a well-timed . . .gS, some­
times supported by . . .h6 and . . . lbh7,
which might secure the eS-square and
imprison White' s light-squared bishop.
Also, the move gS by White is often
no achievement, because Black plays
. . .f6 and activates his rook.
What's left? Let's compare two very
typical plans in the similar main-line
Taimanov Attack structures, but with­
out ...�h4+. One is the familiar pawn
sacrifice eS, and after ...dxeS, fS . Thus
in one fell swoop, White frees his c 1 -
bishop, secures e4 for a knight, ren­
ders Black's g7-bishop 'bad ' , opens
up the tactical possibility of d6, and
even prepares g4-gS ! Against this dan­
gerous plan, Black should be prepared
to make the counter-sacrifice . . . e4, lib­
erating his g7-bishop, negating several
of White's just-mentioned advan­
tages, and gaining eS for his own
knight. In a minority of cases, he may
be able to hold the fort by omitting
. . .e4, defending key squares, and ex­
changing some pieces. For this pur­
pose, a knight on the blockading square
d6 can be useful. In general, the eS/fS
sacrifice is still the biggest worry for
Black in this position, especially with a
queen on f3 . See the examples below. I
92 THE GAMBIT GUIDE TO THE MODERN BENONI
won't pretend that the defence is un­
demanding, but White can easily over­
extend as well.
I had always thought that White was
somewhat better in the comparable
positions in the main Taimanov lines,
primarily because of White's final
plan, which was to transfer a bishop to
g3 (or sometimes h4) by .i.d2-e 1 -g3 or
.i.e3-f2-g3, significantly strengthen­
ing his kingside attack. However, in
our . . .�h4+ lines, the g3-square is al­
ready occupied by a pawn! To my
mind, this is a significant difference,
since g4 has certain drawbacks as
already mentioned. All told, I'm con­
vinced that White has no straightfor­
ward way to launch a successful attack
here, and that Black's chances are
quite as real as White's.
16 a4
As noted above, the dangerous 1 6
e5 ! ? dxe5 1 7 f5 i s possible, and can be
met by the complicated 1 7 ...e4 ! ? (or,
more calmly, 1 7 ... I:tae8 1 8 .i.e4 lLlc4!
=) 1 8 .i.xe4 ! ( 1 8 lLlxe4 lLlxd5 19 .i.g5
f6 ! 20 .i.c4 { 20 .i.d2 lLle5 +} 20 . . .�h8
2 1 .i.xd5 fxg5 and Black's pieces are
active) 1 8 . . .lUe8 ! ? 1 9 J:te1 ( 1 9 .i.c2?!
.i.xc3 20 �xc3 lLlxd5) 19 . . .lLle5 (even
19 . . . .i.xc3 ! ? 20 bxc3 lLle5 can be con­
sidered) 20 �e2 lLlec4 ! ? The . . . e4
idea is not necessarily best here (see
17 . . . J:tae8), but it would appear that
things are dynamically balanced.
The text-move ( 1 6 a4) is Lukacs­
Psakhis, Sarajevo 1 98 1 , which went
16 ... c4 17 �c2 lLlc5 1 8 .i.e3 lLlbd7 1 9
I:tad 1 a6, and now Kapengut's 20
z:i.fe1 ! would give us the same position
as Lau-Dolmatov above (note to Black's
1 5th), except that White has played
a4 instead of a3 and therefore gains
crucial time for his central attack! In­
stead of these hurried solutions with
. . . c4, I think that Black should con­
sider 1 6. . J:tab8 17 a5 lLlc8 intending
. . . a6 and . . . b5, since it discourages the .
idea of a break by e5. For example, 1 8
e5 ( 1 8 .i.e3 b5 ! ? { among others } 1 9
lLlxb5 a6 20 lLlc7 .I:i.xb2+ 2 1 �h 1 J:tb7,
winning the exchange for a pawn, but
with mutual chances) 1 8 . . .dxe5 1 9 f5
lLld6. This is an example of the second
response to the e5/f5 breakthrough
given above. Black' s position looks
solid enough, and he can still contem­
plate returning the pawn by ... e4, or
hunker down in positions like 20 .i.e4
.:!bd8.
In conclusion, I think that the
9 . . .�h4+ variation in response to 9
.i.d3 offers good practical chances. Its
theoretical status is as yet unresolved,
to some extent because 9 .i.d3 is rather
out of favour (9 a4 is played about
seven times as much in my database). I
chose 9 . . .'ilVh4+ because it is consis­
tent and easy to learn. Note that 1 0 g3
'ilVd8 looks like a safer line if you don't
like continuing 1O .. :ii'e7. The reader
should also note the transposition to .
the 9 a4 lines of the next section by
1 2 . . .a6 1 3 a4.
A34)
9 a4 (D)
This line of the Taimanov Benoni i
has caused Black many headaches"
,
and has helped to drive many Benoni :
players away from the 1 d4 lLlf6 2 c4 '
c5 3 d5 e6 4 lLlc3 move-order. White
 PAWN-STORM SYSTEMS
B If, on the other hand, you prefer to
reasons that a4 is necessary in any
case to prevent queenside expansion
by . . . a6 and ...b5. But now he is able to
respond after Black has committed to
a plan; for example, he may want to
put the bishop on e2, d3 or c4 when it
is attacked by, e.g., ... a6 or ...lba6-c7.
This flexibility has made 9 a4 by far
the most popular continuation over the
last 20 years.
Because of the importance of this
variation, I will analyse my recom­
mendations in great detail.
9 :iWh4+
..
Once again, this manoeuvre serves
to weaken White's kingside as a basis
for future counterplay.
But here I should make a very im­
portant point about move-orders. In
the lines I am proposing, Black will
play . . . a6 in the near future, and White
will have a choice between three re­
treats (ignoring .txd7 for the mo­
ment) : .tc4, .td3 and .te2. If, in
response to White's g3, you want to
play the move .. :fie7 versus all three of
these moves, or .. :fid8 versus all three
moves, then the move-order before
93
you is ideal, because it doesn' t hurt to
make that choice now.
play .. :iVd8 against some bishop re­
treats and ... 'JJ!ie7 against others, you
would do well to play 9 ... a6 (D) right
now, to see where White's bishop is
going first.
This also allows you to respond to
one or another bishop retreat by fore­
going . . . 'iVh4+ altogether. A fairly mi­
nor drawback to this strategy is that
after 9 ... a6 1 0 .tmoves �h4+, Black
has committed himself to . . . a6 in
meeting the seldom-seen I I 'it>fl . The
'it>fl move (which is extremely rare
and not even mentioned by some au­
thors) is theoretically easier to meet if
one has the option of playing, e.g.,
... lba6 as well as ... a6. But if you are
satisfied with the 'it>fl lines in this note,
then 9 ... a6 is the most flexible move­
order. Otherwise, see the note to White's
1 0th move (dealing with 9 .. :ilih4+ 1 0
'it>fl ).
There are three 'it>fl lines which
could arise after 9 ... a6:
94 THE GAMBIT GUIDE TO THE MODERN BENONI
a) 1 0 i.c4 'ii'h4+ 1 1 'it'fl fle7 1 2
ttJf3, and now Black can play 1 2... 0-0
1 3 'it'f2 ttJb6 14 i.a2 i.g4, as in the
main lines, or even 1 2. . . i.xc3 1 3 bxc3
'ii'xe4, which is much better than in
similar positions due to the tempo on
c4 and the king on fl ; e.g., 14 'ifb3 0-0
I S 'it'f2 ttJf6 16 l:tel 'ii'fS and White
has less compensation than in similar
lines.
b) 10 i.e2 'ii'h4+ 1 1 'it'fl �e7 1 2
ttJf3 0-0 1 3 eS ! ( 1 3 'it'f2 can b e an­
swered by 1 3 . . .i.xc3 ! ? 1 4 bxc3 ttJf6,
while 1 3 ...ttJf6 is also not bad) 1 3 ...dxeS
1 4 d6 'ii'f6 ! ? (more ambitious than the
safe 14 ...'ii'd S) I S fxeS ( I S ttJe4 'iVfS)
IS . . .ttJxeS 1 6 i.gS 'ii'e6 ! ? ( l 6 ...'ii'fS
1 7 ttJdS ttJbc6) 1 7 ttJdS ttJxf3 ! is un­
clear. Black's idea is I S ttJc7? ttJxgS !
1 9 ttJxe6 ttJxe6 + (or better, consider­
ing the material and White's king­
position).
c) 10 i.d3 is probably the line in
which Black's . . . a6 causes him the
most problems should White play 'it'fl .
Nevertheless, Black should be fine af­
ter 1 O...'ii'h4+ 1 1 'it'f1 'ii'dS ( l l .. .'ii'e7
1 2 ttJf3 0-0 1 3 'it'f2 transposes to note
'bl ' to White's 1 0th move) 1 2 ttJf3 0-0
13 'it'f2 l:teS 14 l:te l 'ii'aS ! I S i.d2 ( 1 S
'ii'd2 ! ? c4 1 6 i.c2 can be answered by
16 ...bS 17 ttJe2 b4 or 1 6. . . ttJcS) I S ...c4
16 i.c2 'ii'b6+ (alternatively, 1 6 ...ttJcS
17 'it'gl i.g4 =) 17 'it'f1 �xb2 I S l:te3
( 1 S l:tbl 'ii'a 3) I S .. :ikb6 19 l:tbl 'iVcs,
and White's compensation looks in­
sufficient.
Let's return to 9 ...'iVh4+ (D):
In what follows, I will analyse both
...'ii'dS and ...'ii'e7 responses. After the
normal 10 g3, I will include ... a6 lines
w
in the analysis, to offer the reader
maximum flexibility. This incidentally
provides a way to construct your own
repertoire. For example, if you play
9 ...a6, there are six different combina­
tions to choose from involving ...'ii'h4+
and ... 'ii'e7 or ... 'iVdS.
Since this is all very confusing,
here's a summary:
1 ) Black can play either 9 ... a6 and
1 O ...'ii'h4+, or 9 . . .'ii'h4+ first.
2) Against both moves, the check
might be answered by 'it'fl , but that is
not much to be feared.
3) The 9 .. :iWh4+ 1 0 g3 'iVretreats
option commits Black's queen to a
single retreat against all white piece
formations, yet leaves open non-... a6
possibilities.
4) The 9 . . . a6 move-order allows
Black to see where the bishop is going .
before he decides upon a retreat for the
queen or perhaps foregoes ...'ii'h4+ al­
together.
S) Having said that, the lines that I
will analyse the most include . . . a6
and could stem from 9 . . . a6. Therefore
I will look at both the moves ...'ii'e7
 PAWN-STORM SYSTEMS
and ...'iVd8 in conjunction with every
bishop retreat from ... a6.
IO g3
If you followed that summary, you
will see that we need to analyse one fi­
nal 'it>fl situation: the lines that don't
transpose to 9 ... a6 1 0 JLretreats 'ieh4+
I I �fl above. Here are a few ideas in
which Black doesn' t transpose to any
of those lines: 1 0 'itfl 0-0 I I lLlf3 'iVe7
(D) and now:
w 16 .:tel 'iVf5 . Moves such as 1 2 ...JLxc3
a) 1 2 JLd3 lLlf6 1 3 h3 ( 1 3 e5 dxe5
{ or 1 3 ...lLle8 = } 1 4 fxe5 lLlg4 1 5 JLg5
'iVxgS ! 16 lLlxgS lLle3+ 17 'itf2 lLlxd 1 +
1 8 l:taxdl lLld7!) 1 3 ...lLla6 ( 1 3 ...lLlh5 !?;
13 ...c4 ! ? 1 4 JLc2 lLlbd7 IS 'iVe2 lLlc5
1 6 �xc4 b6! 1 7 'itgl JLa6 =) 14 'itf2
c4 ! I S JLbl ? ( 1 S JLxc4 lLlxe4+ 1 6
etJxe4 �xe4 =) I s . . .lLlcs 1 6 lLld2 lLlh5
1 7 g3 fS 1 8 l:!e 1 JLxc3 1 9 bxc3 fxe4 20
etJxe4 lLlb3 ! with a large advantage.
b) 1 2 'itf2 and now:
b 1) 12 ... a6 (a good move, although
perhaps not even the best one; note
that this move differs from the analy­
sis above in that Black has played
.. :iVe7) 1 3 JLd3 JLd4+ ! ? 14 lLlxd4 cxd4
95
I S lLle2 ( 1 S lLlb 1 lLlcs 1 6 l:!e I JLg4! 1 7
JLe2 JLxe2 1 8 �xe2 d3 19 �e l d2 ! +)
Is . . .lLlcs 16 b4 ( 1 6 lLlg3 JLg4 =; 1 6
l:!e l JLg4 1 7 'itg l :e8 !) 1 6. . .lLlxe4+
1 7 JLxe4 'iexe4 1 8 'iVxd4, Ward-Hall,
Stockholm 1 988, and now simply
18 .. :iVxd4+ 1 9 lLlxd4 lLld7 intending
. . . lLlf6 or . . .lLlb6 looks at least equal
for Black.
b2) I like the look of the ambitious
pawn-grab 1 2 . . .JLxc3 ! ? 1 3 bxc3 lLlf6
( 1 3 .. :�fxe4 ! ? has to be considered as
well), which looks reasonably safe;
e.g., 14 lLld2 lLlxe4+ I S lLlxe4 'iVxe4
always entail risk, of course, but I
think this casts serious doubt upon 1 2
'itf2.
After these endless 'itfl digressions,
we finally return to 10 g3 (D):
Here Black has a major choice:
A341: IO J!M8 96
•.
A342: IO .'ife7 1 0 1
.•
The differences between these two
moves will become apparent as we pro-
ceed. I tend to prefer 1 O ...'ilVd8 against
i.d3 systems and 1O . . .'iee7 against
96 THE GAMBIT GUIDE TO THE MODERN BENONI
.ie2 systems, but that is by no means
written in stone.
A341 }
1 0 �d8 ll lDf3 0-0 12 0-0
.•.
The untried 1 2 h3 ! ? might be inter­
esting to look at, because it acts to pre­
vent a later ... .ig4:
a) 12 ...J:.e8 1 3 0-0 a6 14 i.d3 ( 1 4
.ie2 lDf6 I S eS ..txh3 1 6 exf6 ..txfl
17 ..txfl .ixf6 looks nice for Black)
14 ... .ixc3 ! ? IS bxc3 lDf6 1 6 l:te l ( 1 6
eS dxeS 1 7 fxeS lDxdS) 1 6. . ..ixh3 1 7
c4 �c8 1 8 'ti'c2 ( 1 8 .ib2? �g4 ! 19
�h2 lDhS +) 18 ... i.g4 1 9 �f2 lDbd7
20 ..tb2 .ihS ! ? with yet another posi­
tion in which White has definite play
for the pawn, but I wouldn't want to
assess this.
b) 1 2. . .lDa6! is a much safer move,
although not as intriguing: 1 3 0-0 lDc7
14 .ie2? ( 1 4 .ic4? lDb6 +; 1 4 .id3
lDf6 I S �g2 a6 1 6 .id2 �b8 17 as
lDd7 ! ? 1 8 �el bS 19 axb6 lDxb6 =,
with moves like . . .fS and . . . .ib7 in
mind) 14 ....ixc3 ! I S bxc3 lDf6 with a
double attack on e4 and h3. This looks
like a very comfortable solution that
exploits the plodding nature of 1 2 h3
by developing quickly.
We now return to 1 2 0-0 (D):
12 a6
.••
This is the recommended move, to
see where the bishop is going before
.ifl becomes possible (i.e., after �el ).
Note that the positions which follow
1 2 . . . a6 arise from 9 ... a6 as well. Two
alternatives of interest are:
a) 1 2 ...lDa6 1 3 .ixd7 .ixd7 (per-
haps 1 3 . . . �xd7 ! ? improves, and if
White continues as in the game with
1 4 fS gxfS I S lDh4, Black has
l S ... fxe4 1 6 lDfS ? { 1 6 lDxe4 fS 1 7
lDgS lDc7 } 1 6 .. :iVxfS ! 1 7 1hfS .ixfS) C
14 fS c4! ? ( l 4... lDb4! ?; 14 ...gxfS I S
lDh4! { I S .igS 'i!Vb6 } l S . . . fxe4 1 6
lDxe4 �e7 { 1 6. . ..ih3 1 7 lDfS } 1 7 %:tel
with quite a serious attack, although
Black might defend by . . . .id4+,
... .ieS, .. .f6 and a timely .. .l::tae8 and
... lDb4) I S .ie3 gxfS ( l S ...lDb4 ! ? is
extremely muddy after 16 lDd2 ;!; or
perhaps 1 6 lDd4) 1 6 .id4! led to mon­
strous complications and a draw in
Si-Soylu, Thessaloniki OL 1 984, but
White is having all the fun here.
b) 12 ....l:te8 1 3 l:[e 1 a6 14 .ifl lDf8 ! ?
I S h 3 lDbd7 ;!; Ziatdinov-Pigusov,
Tashkent 1 986. This prevents most of .
what White wants to do, but Black .
himself has an even harder time find- .
ing a plan.
We now return to the position after·
1 2 ... a6 (D):
We now have (with apologies) a last
split in the material:
A3411: 13 .ic4 97
A3412: 13 .ie2 98
A3413: 13 .id3 100
 PAWN-STORM SYSTEMS
I consider the last move the most
difficult to meet, but they all deserve
close attention.
A34 1 1 )
13 i.c4
This is often given an ' ! ' , but it is
probably not as dangerous as the re­
treat to d3 . White's idea is that when
eS and possibly e6 happen, the bishop
will be ideally placed.
13 liJb6 (D)
...
But this response guarantees Black
an immediate . . .i.g4, developing the
often-cramped queenside and limiting
White's central expansion.
I like the text-move best, but 1 3 .. J�e8
is a decent alternative; then 14 l:!.e l
tiJf8, aimed at preventing eS-e6, is a
little cramped, but Black is better situ­
ated here than in most . . .liJf8 lines;
e.g., I S i.n ( 1 S eS i.g4) IS ...liJbd7 !
(after IS ...i.g4 16 h3 i.xf3 1 7 'iVxf3
ctJbd7 Black's position is ideal except
for the knight on f8, which needs to be
on the queenside, so White should
maintain a small but persistent edge)
16 h3 l:tb8 17 c.t>h2 Wilc7 18 .l:.e3 ! ? c4
97
1 9 as bS 20 axb6 liJxb6 2 1 l'le2 liJfd7
22 liJd4 i.b7 = with, e.g., ... liJcs to
follow, H.Sl'lrensen-V .Peicheva, Co­
penhagen 1990. Several strong Benoni
players have taken up this ...liJf8 idea,
which is appropriate under some cir­
cumstances, but also runs the risk of
prolonged passivity.
14 i.a2
It seems slightly inconsistent to
play 1 4 i.e2 after 1 3 i.c4, but White
hopes to defend the kingside and at
some point play aS to establish a space
advantage on both sides of the board.
In fact, 1 4 i.e2 led to a difficult game
after 1 4 ... i.g4 I S liJgS (harmless are
I S i.e3 �e8 and I S �el liJ8d7 1 6 as
i.xf3 { 1 6. . .liJc8 ) 1 7 i.xf3 liJc4 1 8
'iVa4 bS 1 9 axb6 liJcxb6 20 'iVc2 :e8)
I S . . .i.xe2 16 �xe2 Wie7 17 aS liJc8 1 8
i.d2 liJd7 1 9 %hel in H.Olafsson­
Psakhis, Moscow 1 989. This position
is given as ;t in the books, but Hebert
points out an improvement on the
game, which itself was drawn: 1 9 ... bS
20 axb6 and instead of 20 . . . i.d4+,
which did not turn out so badly, Hebert
98 THE GAMBIT GUIDE TO THE MODERN BENONI
suggests 20 ...liJcxb6 2 1 eS?! (21 b3 c4
22 b4 as ! and 2 1 :tal h6 22 liJf3 c4 ! ?
23 l:.fe l :tfe8 are equal) 2 l .. .dxeS 22
fS liJf6 ! and as Hebert says, "There is
no breakthrough in sight for White".
Apparently true, since 23 liJge4 can be
met by 23 ...gxfS .
14 �g4 1S aS liJc8!?
..•
We've seen this idea before: the
b8-knight goes to d7, and the c8-
knight can assist in a queenside ad­
vance. Also possible is just l S . . . �xf3
1 6 'ii'xf3 liJc8 1 7 �b3 liJd7 =.
16 1Wb3
Attacking b7 and unpinning the
knight at the same time. Otherwise
White will have a difficult time expel­
ling the bishop from g4.
16 :ii'xaS
••
1 6 ...:ta7 ! ? may be playable, but
seems unnecessary.
17 'ii'xb7 liJb6 18 i.d2 �xf3 19
:txf3 liJ8d7 =
Black intends ... :tab8 or ...:tfb8
and, in the right position, ... liJc4.
A34 1 2)
13 �e2
It is unclear whether it is easier to
play against this in the lines with
1O . . .1We7 or in those with 1O ...'i'd8.
White would just have a favourable
version of the main line of the Four
Pawns Attack if it weren't for that irri­
tating weakness created by g3. As it
stands, Black has full-fledged play.
13 %:te8 (D)
.••
This is probably best, anticipating
e5-e6 after ...liJf6 and allowing ... liJf8
in some lines. 1 3 ...liJf6 seems worse
due to the rarely-played 1 4 eS ! (most
critical; 14 'i.t;lg2 i.g4 1 5 liJd2 i.xe2
16 'iYxe2 liJbd7 17 liJc4 liJb6 1 8 %:te 1
%:te8 1 9 1Wd3 liJxc4 20 'i'xc4 'iYaS ! 2 1
i.d2 'ii'b4 = Rivas-Reinaldo Castineira,
Burgas 1 999) 14 ... liJe8 I S %:te l (White
is better here; compare the same posi­
tion in Line A3422 where ...'iYe7 is al­
ready in) I S ...liJc7 ! ? 16 i.e3 ( 1 6 exd6
liJe8 ! ? 1 7 i.e3 liJd7 1 8 liJe4 i.xb2 1 9
%:tb l ;1;) 1 6... :te8 1 7 liJe4 ! dxeS 1 8 d6
liJe6 1 9 fxeS liJd7 20 'ii'd5 ;t. These
lines demonstrate the dangers of ...liJf6.
If Black can successfully get ...i.g4
and ... i.xf3 in, he will usually equal­
ize, but he has to watch out for eS .
Hence the text-move, 1 3 ...%:te8.
w
Now (after 1 3 ...%:te8):
a) 14 liJd2 liJf6 (or 14 ...i.d4+ 1 5
'i.t;lg2 liJf6 1 6 i.f3 'iYd7 1 7 'i.t;lh l 'iYe7 ! =
with . . . hS next, Geers-Betker, COIT.
1 990) I S i.f3 i.h3 ! 1 6 %:tel liJbd7
( 1 6...liJfd7 17 liJc4 liJb6 is also logi­
cal) 17 liJc4 liJb6 1 8 liJe3 hS 19 'i'd3
%:tb8 20 i.d2, Hulak-Lobron, Zagreb
1 985, and now Kharitonov suggests
that 20 ...liJbd7 gives Black an equal
game.
 PAWN-STORM SYSTEMS
b) After 14 .:tel , 14 ... iLxc3 1 5 bxc3
i:txe4 should be strongly considered;
e.g., 1 6 c4! 'Llf6 17 iLb2 and then:
b1) 17 . . . iLg4 1 8 h3 iLxf3 1 9 iLxf3
Itxe1+ ( 1 9 .. Jhc4 ! ? 20 �d3 .:tb4 2 1
iLc3 'Llbd7 22 iLxb4 cxb4 with enough
for the exchange, Koskinen-Danner,
COIT. 1 985) 20 'iWxe 1 'Llbd7 2 1 g4 'iVb6
22 l:tb1 'ii'b3 23 'iWc3 'ii'xc3 24 iLxc3
i:.b8 = Arkhipov-Sax, Hungary 1 984
(material versus two bishops).
b2) Another logical idea is 1 7 ....:te8
1 8 'iWb3 iLg4! 1 9 'iWc3 ( 1 9 iLfl 'Llbd7
20 �c3 'iWb6 2 1 'Lld2 'iWb4; 1 9 'iWxb7
'Llbd7 20 'iib 3 l:lb8 2 1 'iVc2 'iWb6)
1 9 . . .'Llbd7 + with the idea ...'iWb6-b4
(with ...b6 versus a5), or in some cases
doubling on the e-file. I think that
White is falling short here.
c) 14 'it>g2 is the normal move:
14 . . . 'Llf6 ( 14... iLxc3 ! ? 1 5 bxc3 ':xe4
16 iLd3 l'te8 17 c4 'Llf6 1 8 h3 'Llbd7
19 iLb2 :b8, with the usual and very
real risks to Black, nevertheless led to
a win for her in S.Clausen-E.Peicheva,
Copenhagen 1 989) and now:
c l ) After 1 5 'Lld2, 1 5 . . .'Llbd7 1 6
iLf3 l:tb8 equalized i n Muir-Prasad,
London 1 987. One might also want to
look into 1 5 . . .'iVd7 ! ?; e.g., 1 6 f5 'iWc7
1 7 fxg6 hxg6 1 8 'ii'c2 'Llbd7 1 9 'Llc4
'Llb6 20 'Llxb6 'i¥xb6 2 1 a5 �c7 22
iLg5 'Lld7 =.
c2) 15 e5 dxe5 16 fxe5 'Llg4 1 7
iLg5 ! ? ( 1 7 e6 fxe6 1 8 'Llg5, Muir-E.Pei­
cheva, Copenhagen 1 990, 1 8 ... 'Lle5 1 9
dxe6 { 19 iLf4 b6 } 1 9 . . :it'xd1 20 l:txd 1
b6! =) 1 7 ... 'iVxg5 ! 1 8 'Llxg5 'Lle3+ 1 9
'it>h 1 'Llxd 1 20 l'taxd 1 iLxe5 2 1 iLc4!
(2 1 %:td2 f6; 2 1 d6 gives Black the extra
possibility 2 1 .. .iLxc3 22 iLc4 'it>g7 !)
99
2 1 .. .f6 22 d6+ 'it>g7 23 'Lld5 ! iLg4! 24
'Llc7 lLlc6! 25 'Llge6+ (25 'Llxa8 1ha8
26 'Lle6+ comes to the same thing)
25 . . .l:txe6! 26 'Llxe6+ iLxe6 27 iLxe6
J:.d8 28 d7 'Lld4 29 iLh3 a5 = 30 l:tfe1 ! ?
g5 ! and ...h5.
d) 1 4 'iVc2 is a sensible move,
though the queen may tum out not to be
particularly well placed on c2. Then:
d 1 ) 14 ... b6 is possible, contem­
plating . . .l:ta7 and/or . . .'Llf6.
d2) 14 ...'Llf8 intending ...iLg4 might
exploit White's queen position; e.g.,
1 5 e5? ! dxe5 16 fxe5 iLf5 1 7 iLd3
iLxd3 1 8 'ikxd3 'Llbd7 + 1.BjeITe­
E.Peicheva, Vejstrup (4) 1 989. In­
stead, 1 5 f5 'Llbd7 is equal, while 1 5
'it>g2 iLg4 followed by . . .'Llbd7 i s a fa­
miliar idea: Black can advance on the
queens ide and place his knights on d7
and b6 or f6 as needed.
d3) 14 ...'Llf6 1 5 e5 ! ? dxe5 1 6 fxe5
'Llg4! ( 1 6 ...'Llxd5? ! 1 7 'Llxd5 'iVxd5 1 8
iLc4 !) 1 7 iLg5 'ikc7 1 8 d6 ! ? ( 1 8 e6
fxe6 1 9 .:tad1 c4 is double-edged)
1 8 .. :�c6 is unclear; e.g., 1 9 'Llb5 ! ( 1 9
iLc4 iLe6; 1 9 iLb5 axb5 20 axb5 J:.xa1
2 1 bxc6 J:.xfl + 22 'it>xfl 'Llxc6 with
compensation) 19 . . .iLf5 20 'ikd2 (20
'ii'b 3? axb5 2 1 iLxb5 'iib6) 20. . .'Lld7
2 1 'Llc7 'Llgxe5 22 'Llxa8 (22 'Llh4
iLe4 23 'Llxa8 l:txa8 = 24 b4! ? cxb4 25
�fc 1 'iVb6+ intending 26 iLe3? iLh6!
-+) 22 ... .:txa8 and now:
d3 1 ) 23 'Llh4 iLe6 ! ? (23 . . .iLe4) 24
a5 b5 ! 25 axb6 'Llxb6 26 l:txa6 (26
iLxa6 'Llec4) 26 . . .l:txa6 27 iLxa6
'Llec4 28 iLxc4 'Llxc4 +.
d32) 23 'Llxe5 'Llxe5 24 iLh6 iLh8
25 a5 (25 l:tac 1 iLh3 26 ':f2 %1d8 27
'ike3 b6 28 a5 iLc8 ! ? 29 iLfl 'iVxd6 30
100 THE GAMBIT GUIDE TO THE MODERN BENONI
�f4 lLlg4 !) 25 . . . �e4 (or 25 . . . �h3 26
l:.f2 �g4 ! ? 27 �f4 �xe2 28 �xe2
�xd6 29 l:td l 'fie7 = ) 26 l:.adl c4 27
�f4 lLld3 28 �xd3 �xd3 29 l:tf2 l:te8
This is not for the faint of heart.
As 14 'fic2 isn't particularly pointed,
one should also consider Black's 14th
move alternatives.
A34 13)
13 �d3 lLlf6 (D)
w and space count for something, but I
This gives Black a shot at . . . �g4 or
. . . �h3, since 1 4 e5 loses the d-pawn.
There are three logical answers, by far
the most fascinating and dangerous of
which is ...
14 �b3!
Other moves:
a) 14 f5 lLlbd7 1 5 �g5 'fic7 1 6
lLld2 l:tb8 1 7 a5 h 6 (Hebert analyses
1 7 ... lLle5 1 8 lLlc4 lLlxc4 1 9 �xc4 b5
20 axb6 'fixb6 and now 21 lla2 gxf5 ! ?
22 exf5 l:te8 or 2 1 'fid3 lLlg4 2 2 f6
�h8 =+=) 1 8 �f4 g5 1 9 �e3 lLle5 20
lLlc4 lLlfg4 ! 2 1 �c l lLlxc4 22 �xc4
�d4+ 23 �g2 lLle5 24 �h5 f6 ! ? 25
�e2 �g7 =+= Ragnarsson-H.Olafsson,
Reykjavik 1 998. Perhaps not a model
game, but I particularly like the way in
which the e5-square eventually turned
the play in Black' s favour.
b) 14 �g2 �g4 (this bishop might
not get out if White is allowed to play
h3 after, for example, 14 . . .lLlbd7 or
14 .. ":ie8) 15 h3 �xf3+ 16 'ii'xf3 lLlbd7
reaches a standard position of a type
we've seen before. Black's pieces de­
velop smoothly, he has restraint on e5
and he can expand on the queenside.
White has a hard time finding a good
plan. Of course, White's two bishops
think that Black has equalized here. 1 7
�e3 ( 1 7 e5 dxe5 1 8 f5 e4! 1 9 �xe4
lLlxe4 20 'fixe4 �xc3 2 1 bxc3 lLlf6 =+
Hebert; 1 7 �d2 l:tc8) 1 7 . . .'iVa5 1 8
::rael Xtac8 1 9 �gl lLle8 20 �c2 l:tb8
2 1 'ue2 b5 22 axb5 axb5 23 l::.fel c4 24
e5 dxe5 25 fxe5 b4 26 lLldl lLlc7 =+=
P.Schwarz-Pigusov, Biel 1 989.
14...�h3!
Or:
a) 14 . . .lLle8 1 5 e5 ! ? �e7 ;j; 1 6 %tel
.i.g4 17 exd6?? .i.d4+ ! .
b ) 14 . . .'iYe7 i s not very attractive,
but possible, intending 1 5 e5 lLle8 or
1 5 .l:.e 1 lLle8.
15 l:tel lLlg4! (D)
A truly incredible solution, leaving
pawns and pieces hanging. Now:
a) Obviously not 1 6 lLlg5?? �d4+.
b) 1 6 'iYxb7 lLld7 17 a5 ( 1 7 'iYc6
�d4+ 1 8 lLlxd4 cxd4 1 9 'iVxd6 dxc3
20 bxc3 :tc8; for example, 2 1 �xa6
l:txc3 22 �b4 { 22 �b2 'fia8 ! wins for
Black } 22 . . .'iYf6 =+= 23 �e2? l:tc4! 24
e5 lLldxe5 -+) 1 7 ... 'iVe7 and then:
 PAWN-STORM SYSTEMS
b I) 1 8 1i.xa6 !tab8 and now 1 9
'VJIic7?? tDdeS ! - + was played i n the
exciting game Pigusov-Garcia Marti­
nez, Moscow 1 987. White's 1 9th was
disastrous, but he would also have
done badly after 1 9 'VJIic6 l:tb4 20 1i.d3
(20 1i.e2? tDb8 2 1 "iWa8 !'txe4 ! 22
tDxe4 'iVxe4 23 �h l tDf2+! 24 �gl
CDd3 2S l:.a4 c4 -+; 20 !'te2? tDb8 2 1
'VJIia8 CDxa6 22 'iiVxa6 1i.xc3 2 3 bxc3
�xe4 and wins) 20. . .tDb8 2 1 'iiVa8 l:tb7
(2 l . ..CDd7 22 'ili'c6 tDb8 repeats, but
Black is right to play for more than
this) 22 tDbS 'ili'd7 ! and now 23 tDgS?!
�xbS 24 1i.xbS (or else . . .l:tb7-a7)
24 . . .1i.d4+ 2S �h l tDf2+ 26 �gl
'iVxbS =1= or 23 a6 !txbS 24 a7 ! !tb6 =1=
(or 24 ... l:tb4 =1=); e.g., 2S axb8'iV l:tfxb8
26 'ili'a2 l:tb3 27 l:td l c4 ! .
b2) Psakhis's suggested alternative
1 8 'ili'b3 can be met by a plethora of
tactics: 1 8 ... c4 ! 1 9 1i.xc4 ( 1 9 'ili'xc4
CDcS and ...�a7 next) 1 9 ...tDcS 20 'ili'c2
.l:.fc8 (20 . . .'ika7 ! ? 2 1 �h l 1i.xc3 22
bxc3 CDxe4 ! 23 tDd4 l:tfe8 gives Black
counterplay) 2 1 1i.fl (21 !ta2 "Wia7 22
'it>h l CDd3 ! 23 1i.xd3 tDf2+ 24 �gl
CDxd3+ 2S lIe3 CDb4 -+; 2 1 1i.d2? !
101
'iiVa7 =+= 22 �h l CDd3 ! 23 .itxd3 'iiVf2
-+; 2 1 l:ta3 'Wia7 22 .i.e3 CDxe3 23
l:.xe3 CDb7 ! ? 24 CDd l �cS 2S tDf2 !
tDxaS ! =1=) 2 l . ..�a7 22 �h l 1i.xfl 23
�xfl l:.ab8; e.g., 24 CDgS tDb3 2S l:ta3
tDxc 1 26 11xc 1 'iVf2! =1=.
c) In lnformator, Pigusov suggests
1 6 1i.fl '±' . But this once again lands
us in a swirl of complications (in
which Black is at least holding his
own) after 16 ... 1i.xfl :
c l ) 17 �xfl 'iVc8 1 8 h3 tDh2+ ! 1 9
�g2 tDxf3 20 �xf3 �xh3 2 1 'ili'xb7
CDd7 22 .i.d2 gS ! =1=.
c2) 1 7 ':'xfl 'iVc8 ! ? 18 h3 tDf6 19
eS CDe8 =; e.g., 20 g4 tDd7 2 1 e6 c4 22
'iVc2 fxe6 23 dxe6 tDcs 24 tDdS 'iVxe6
2S �xc4 tDc7 ! =1=.
c3) 17 "Wixb7 tDd7 18 l:txfl ':'e8
( 1 8 ... 1i.d4+ 1 9 tDxd4 cxd4 20 CDe2
l:te8 { or 20 ... 'iVe7 ! ? 2 1 tDxd4 �xe4 = }
2 1 tDxd4 CDcS 22 'iVb4 l:tb8 2 3 "iWa3
tDxe4 24 CDc6 'iiVb6+ 2S �g2 ':'bc8
with equal chances) 1 9 h3 .i.d4+ 20
�g2 (20 tDxd4 cxd4 2 1 hxg4 CDcS)
20 ... CDgf6 2 1 eS dxeS 22 fxeS tDxeS 23
CDxeS 1i.xeS =1=.
I have annotated this game in detail
to illustrate how dynamic and enter­
taining the play can be in the Benoni,
especially when Black invests mate­
rial for attack. The lesson, as so often,
is that one must remain active in or­
der to counteract and exploit White's
greater control of terrain. This sort of
game can teach you more about the
Benoni than pages of listed variations
might.
A342}
10 :�e7 (D)
..
102 THE GAMBIT GUIDE TO THE MODERN BENONI
w 22 'i!Vxd3 'i!Vxh3+ 23 'ii;>e2 'iVg2+ 24
When Black makes this critical de-
cision, he is counting upon certain ad­
vantages that it has over l O . . .'i!VdS, i.e.,
more pressure on the e-pawn and
piece development that clears the bank
rank. On the negative side, the queen
can't swing to squares like b6 and as.
In addition, if White plays eS and this
is met by . . . dxeS, then d6 comes with
tempo, and in those cases where
White does not play d6, then White is
helped by the fact that there is no di­
rect attack on White's dS-pawn. The
reality is that both moves are playable,
and that each player will find different
contexts in which one is to be pre­
ferred over the other. I should again re­
mind the reader that by playing the
9 . . . a6 move-order, one can play either
move against a given bishop retreat.
11 tDf3
Or:
a) 1 1 iVe2?! 0-0 1 2 tDf3 commits
the queen to a square it may very well
not want to go to. A nice example is
1 2 . . ..l:!.eS 1 3 .id3 tDa6 1 4 0-0 tDb4 I S
.:I.el tDf6 1 6 'ii;>g2 .ig4 1 7 h 3 .ixf3+
I S ..t>xf3 a6 1 9 g4?! hS ! 20 gS ? 'ikd7 !
2 1 �fl (2 1 gxf6 iVxh3+ 22 'ii;>f2
tDxd3+; 2 1 'ii;>g2 tDfxdS +) 2 l . . .tDxd3
'ii;>d l tDg4 2S �d2 tDf2+ 26 'ii;>c2 .ixc3
with . . .l'.!e4 to come, O.Garcia-Ham­
douchi, Sitges 1 994.
b) I I 'iVf3 tDa6 12 tDge2 tDb4 1 3
0-0 0-0 1 4 g4 and now Black can play
14 . . . tDc2 + with . . . tDd4 next, as in
Hort-Hulak, Surakarta 1 982. Kapen­
gut gives 1 4 . . . a6 I S .ic4 tDb6 1 6 b3
tDc2 1 7 l'.!a2 tDd4 1 8 'ifg2 and now
I S .. .fS 'unclear' , but instead I S ...tDxc4
1 9 bxc4 tDxe2+ ! 20 tDxe2 J:IeS is aw-
ful for White.
1 1 0-0
•.•
1 l .. ..ixc3+ 1 2 bxc3 'ikxe4+ 1 3 'ii;>f2
0-0 has also been played, with the
usual risks. When the queen (instead
of a rook) has to capture on e4 and
there is no gain of tempo on a bishop
at c4, those risks are greater.
12 0-0 a6 (D)
With this particular move-order,
1 2 . . .tDa6 with the idea . . .tDb4 is also
possible, but as explained in the last
section, . . . a6 may already have been
played.
w
 PAWN-STORM SYSTEMS
Now we have the same split that we
saw after 1 O. . .'iUd8, Le.:
A3421: 13 i.c4 1 03
A3422: 13 i.e2 1 03
A3423: 13 i.d3 1 05
A342 1 )
13 i.c4
This is probably the easiest move to
meet, although I lack examples.
13 lbb6 14 i.a2
•••
14 i.e2 is met by 14 ... i.g4 15 a5
lbc8. Compare Line A34 1 1 (the anal­
ogous line with 1 0 .. :iM8); Black is
better off with his queen developed.
14 i.g4 15 a5
••.
Instead, 1 5 e5 dxe5 1 6 d6 'iVd8 l 7
fxe5 ..txe5 doesn't do much for White.
Nor does 1 5 'ii'b 3 ! ? ..txf3 1 6 ':xf3
lbc8 1 7 i.d2 lbd7 ! 1 8 'ii'xb7? ( 1 8 Wg2
l:tb8 =) 1 8 . . . lbe5 ! and Black has a
clear advantage.
15 lbc8 16 l:tel
•••
After 16 �d3 i.xf3 ! ? 17 'iVxf3 lbd7,
the traditional pawn sacrifice 1 8 e5? !
dxe5 1 9 f5 fails to 1 9. . .lbd6, which is
generally the case with a knight on c8.
Now (after 1 6 l:te l ) a plausible fol­
low-up would be 16 . . . lbd7 17 h3
.i.xf3 18 'iVxf3 l:.b8 ( 1 8 ... �d8 ! ?) 1 9
.ib3 ( 1 9 e5 dxe5 20 f5 lbd6; 1 9 i.c4
J:te8) 1 9 ... l:te8 20 i.d2 b5 2 1 axb6
Itxb6 =.
A3422)
13 i.e2
This is a complicated line. White
lends less support to e4 and blocks the
e-file, but he anticipates ... i.g4 and
keeps d5 protected. As we shall see,
this makes it easier to play e5.
103
13 lbf6
•.•
1 3 .. .l::t e8 ! ? is quite interesting; for
example, 14 �g2 ( 1 4 ':el lbf8 intend­
ing . . . i.g4 and ... lbbd7 is unclear)
14 . . .i.xc3 ! ? 1 5 bxc3 'iUxe4 1 6 ':e l
can be met by 1 6. . .'ii'f5 or 1 6 ...lbf6 l 7
i.xa6 lbxa6 1 8 ':xe4 lbxe4 1 9 c4 lbb4
20 i.b2 i.h3+ 2 1 �gl ..tg4 with
promising compensation.
14 e5 lbe8 (D)
W
This is critical for the ...�e7 block­
ing strategy. Black hopes to destroy
the centre by . . . i.g4 and . . . lbd7, but
White has the dangerous e5-e6 idea,
which wasn't available in most lines
after .. :�d8 because White's d-pawn
was loose.
15 e6
This is the most intimidating move.
Other ideas:
a) 1 5 lIel .ig4 (the most ambi­
tious; after 1 5 . . . lbd7, 1 6 exd6 ! ? �d8 !
1 7 a5 lbxd6 was roughly equal in
Grabuzova-Balashov, Moscow 1 996,
but 16 e6 is probably more important)
1 6 e6 ( 1 6 i.c4 transposes to line 'b' )
16 . . .fxe6 (I think this must be better
104 THE GAMBIT GUIDE TO THE MODERN BENONI
than 1 6. . . .ltxf3, as played previously)
and now:
a l ) 17 dxe6 lbc7 ! 1 8 .ltc4 lbc6 +:
intending ... lbd4.
a2) 17 lbg5 ! .ltxe2 1 8 �xe2 ( 1 8
l:txe2 .ltd4+ 1 9 .lte3 e5) 1 8 . . ..ltd4+ 1 9
�h l ( 1 9 .lte3 e5 20 lbe6 .ltxe3+ 2 1
'iVxe3 ':'f5 !) 1 9 ...e5 20 lbe6 ':'f7 2 1
lbg5 ':'f8 22 lbe6 ':'f7 = 23 lbe4 ! ? lbg7
24 lb4g5 lbxe6 and now 25 lbxf7?
"fixf7 26 dxe6 �xe6 gives Black the
advantage.
b) 1 5 .ltc4 .ltg4 1 6 ':'el dxe5 1 7
fxe5 .ltxf3 1 8 �xf3 lbd7 and White's
centre is vulnerable: 19 .i.f4 ( 1 9 d6
lbxd6 ! ; 1 9 e6 lbe5 20 �e2 lbd6 +:)
19 ...lbxe5 20 �e2 lbd6 2 1 .ltxa6 ':'fe8 !
=t .
15 fxe6 16 .ltc4
••.
1 6 ':'e l is met by 1 6. . . lbc7.
16 lbc7
•••
16 ...�h8 ! ? is another way to play
it, intending 1 7 ':'e 1 ( 1 7 dxe6 lbc7)
17 ... e5 1 8 fxe5 dxe5 1 9 .ltg5 �c7 with
the ideas . . ..ltg4 and ... lbd6.
17 l:.el
Here we see one advantage of
10 .. :ii'e7 over 1O . . .�d8 in this kind of
position: the e-pawn is blocked, so
Black has some time to organize his
minor pieces. In Petursson-J.Fries
Nielsen, Nrestved 1 988, Black now
played the rather panicky 1 7 . . .b5? ! ,
giving back the pawn without fully re­
leasing the pressure. It seems to me
that Black should first get his king out
of the way by. . .
17. .�h8!? (D)
. change.
This helps the idea of ... e5 in sev­
eral positions (and avoids nasty shots
like f5, meeting . . Jhf5 with .i.g5 !).
The position defies complete analysis,
but here are some ideas:
a) 1 8 dxe6 lbc6 1 9 lbd5 ( l 9 lbg5
lbd4 { there are alternatives here } 20
lbf7+ ':'xf7 ! 2 1 exf7 lbf3+ 22 �xf3
�xel + 23 �g2 .lte6 24 .i.xe6 'ilVxe6
25 �xb7 �xf7 = ) 1 9... lbxd5 20 'iIIxd5
(20 .i.xd5 lbb4 2 1 .ltb3 d5 ! 22 lbg5
c4) 20. . .:f5 (20...lbb4 2 1 �e4 b5 ! is
interesting; e.g., 22 �xa8 bxc4 23 f5
.ltb7 24 �a7 ':'xf5 25 .ltg5 lbc6 26
'iIIb6 ':'xg5 27 lbxg5 .ltd4+ 28 �g2
�xg5 29 �xb7 �d5+ 30 �h3 'iIIh5+
=) 21 �e4 lbb4 22 lbh4 (22 g4 d5 23
�e2 ':'f6 -+) 22 ... d5 ! 23 lbxf5 gxf5
24 �xf5 dxc4 and Black's material
may well outweigh White's bind.
b) 1 8 .lte3 exd5 ( l 8 ...lbd7 19 dxe6
lbb6 is unclear) 1 9 lbxd5 ( 1 9 .ltxc5
�d7 -+) 1 9 ...lbxd5 20 'iVxd5 'ilVc7 2 1
lbg5 lbc6! 2 2 lbf7+ �xf7 2 3 �xf7
:txf7 24 .ltxf7 ..txb2 25 :tabl .i.c3 26
':'edl lbd4 =. Black has the bishop­
pair, a pawn and activity for the ex­
c) 1 8 lbg5 e5 ! ? (the alternative
1 8 ... ..td4+ 1 9 �h 1 e5 is worth consid­
ering) 1 9 fxe5 ..txe5 20 .ltf4 lbd7 2 1
 PAWN-STORM SYSTEMS
CDf3 ! (2 1 CDe6 CDxe6 22 dxe6 �d4+
23 �h l CDe5 +) 2 l ...CDe8 ! ? 22 �e2 b6
(22 . . .�g7 ! ?) 23 �h6 (23 CDxe5 CDxe5
24 �xe5+ dxe5 25 �xe5+ �xe5 26
':'xeS CDd6 = ) 23 . . ..:.f7 (23 . . ..:.fS ! ? 24
�d3 ':'hS) 24 CDg5 %:.fS 2S CDe6 CDdf6
with the point that 26 g4 is met by
26 . . .�xe6 ! .
Black has to defend carefully here.
With best play for both sides, I think
one arrives at a kind of dynamic equal­
ity in which, as we have just seen, it is
as easy for White to go wrong as it is
for Black. Still, it has to be admitted
that such a perilous course will not ap­
peal to everyone. Less adventurous
players may prefer 1 3 ...%:.e8, or else to
look into the lines with ...�d8.
A3423)
13 i.d3 CDf6 (D)
Other ideas seem to commit Black
to the ...tiJf8 retreat:
a) 1 3 ...b6! ? 14 l:.el :e8 I S �c2
l1a7 1 6 b3 CDf8 1 7 �b2 i.g4 was about
equal in Vrbata-Pekarek, Czech Cht
1995/6. Of course, White has many al­
ternatives.
b) 13 ... ':'e8 14 l:.el CDf8 with the
same idea of . . .i.g4 is solid, but this is
also cramped.
14 �g2
Preparing to meet ... i.g4 with h3.
As usual, White has some critical al­
ternatives here:
a) 14 l:.e l i.g4 I S �f1 CDbd7 1 6
h3 i.xf3 1 7 �xf3 tiJe8 resembles the
... �g4 Iine against the Four Pawns At­
tack: 1 8 �e3 �d8 (to head to the
queenside, as in a 1 O...�d8 variation)
19 g4 l:.b8 (perhaps 1 9 . . .'iWaS ; for
105
example, 20 �f2 %:.b8 2 1 �g2 c4 22
lIed l tiJcs =) 20 �f2 CDc7 2 1 �g3
lIe8 22 lIe2 (22 as bS 23 axb6 lhb6
24 ':'a2 �8 25 lie2 lIb3 ! =) 22...bS
23 lIae l , Fronczek-Namgilov, Kato­
wice 1 99 1 . Now the standard line­
opening idea 23 ...bxa4 ! looks effec­
tive: 24 eS (24 CDxa4 CDb5 threatens
...CDd4) 24. . .CDbS 2S CDxb5 ! ? axbS 26
e6 �d4+ 27 �h 1 fxe6 28 dxe6 CDf6
29 f5 gxfS 30 gxfS �h8 with ongoing
complications that are difficult to as­
sess.
b) 1 4 eS CDe8 ( l 4 . . .tiJfd7 ! ?) I S
exd6 (after I S lie! i.g4, 1 6 e6 fxe6 1 7
dxe6 CDc7 1 8 �c4 transposes to note
' a i ' to White's l Sth move in Line
A3422, while 16 �c4 is line 'b' of that
note) 1 5 ...CDxd6 1 6 lIel �c7 1 7 tiJeS
tiJd7 1 8 tiJc4 CDf6 1 9 �b3 CDxc4 20
�xc4 i.h3 ! 2 1 �e3 lIac8 22 :tad 1
CDg4 =.
c) Kapengut likes 14 �b3, which
unpins White's f3-knight, but doesn't
threaten b7 and therefore compares
rather poorly with 1 4 �b3 in Line
A34 1 3 . Black should play 14 ...�h3,
leading to:
106 THE GAMBIT GUIDE TO THE MODERN BENONI
c 1 ) I S ':'f2 tZJg4 1 6 l:te2 tZJd7 1 7
'iixb7 1:1fb8 1 8 'iic7 1:Ic8 1 9 'iiaS c4 20
�c2 tZJdeS (20 ...tZJcS ! ? may be better)
2 1 fxeS 'iia7+ 22 'iiih i tZJf2+ 23 'iiig i
= Black got lucky in an inferior endgame
c2) I S ':'e 1 tZJg4 (anyway !) 1 6 eS
( 1 6 �d2 can be met by 16 ...tZJd7, hop­
ing for 17 'iixb7?? tZJdeS !) 1 6...dxeS 1 7
tZJgS c4 ! 1 8 'iixc4 ( 1 8 .Jixc4? 'iicS+)
18 ...bS 19 'iie4 'fia7+ 20 �e3 tZJxe3
2 1 'fixe3 'fixe3+ 22 1:Ixe3 exf4 23 1:If3
b4 and Black stands better.
14 .Jig4 15 h3 .Jixf3+ 16 'iixf3
••.
tZJbd7 (D)
A familiar type of situation, dis­
cussed at length in Line A33 (note to
Black's l Sth move). White is a long
way from a breakthrough, and one
question is whether Black's queenside
play will be well underway by the time
his opponent works up real threats.
17 g4
1 7 .Jid2 is a solid move, but hasn't
established any advantage in two tries,
which strengthens one's belief in
Black's set-up: 1 7 ...1:Ifc8 (Black was
able to cope with top opposition even
after the less logical 17 .. Jlac8? ! : 1 8 b3
tZJe8 1 9 ':'ac 1 tZJc7 20 as oUb8 2 1 1:Ife 1
bS 22 b4! ?, Yusupov-Villwock, Essen
2000; this is probably about equal, but
and actually won) 1 8 ':'ae l c4 1 9 .Jib l
tZJe8 20 g4 bS ! with timely counterplay,
Sergienko-Sziebert, Budapest 1 993.
The text-move ( 1 7 g4) is Kreuzer­
Aulinger, Germany tt 1 996/7, which
continued 17 ...OUfe8 ! ? 1 8 gS tZJhS 1 9
�d2 ::tab8, which appears somewhat
in White's favour. However, I don't see
anything wrong with either 1 7 . . .h6,
with the idea 18 h4 hS ! 19 gS tZJg4 20
fS �d4 =, or 17 .. J:tfc8 1 8 gS tZJe8.
Overall, I tend to like 1O .. :tWd8 in
more situations than 1O ... 'iie 7, as it
avoids some of these scary eS lines;
but sometimes the latter move is use­
ful, and by choosing 9 ... a6, one can
play either according to one's taste.
B)
7 'fie7!? (D)
•.•
w
An eccentric idea which avoids all
of the sections above by deviating at
 PAWN-STORM SYSTEMS
the earliest possible point. The main
incentive for playing this way would
be to avoid the Taimanov Attack.
S lLlf3
Direct and probably best, since
8 ... lLlxe4?? now fails to 9 'iVa4+. Other
moves have been tried:
a) 8 i.d3 i.g7 9 lLlf3 0-0 1 0 0-0
.Jtg4 transposes to note 'b' to White's
9th move.
b) 8 'iVc2 i.g7 9 i.e2 0-0 1 0 lLlf3,
Watarai-Taleb, Asian Cht 1 987, and
now 1O ...lLlxe4! 1 1 lLlxe4 i.fS is strong:
1 2 lLlfgS l:.e8 1 3 i.f3 i.xe4 14 i.xe4
h6 ( 1 4. . . fS l S lLle6 fxe4 1 6 0-0 lLlc6 ! )
1 S lLle6! lLlc6 ! ? ( 1 S . . . fxe6) 1 6 0-0 ( 1 6
"iDe2 lLld4 1 7 lLlxd4 �xe4) 1 6. . . lLld4 +.
c) 8 i.bS+ i.d7 ! ? (8 ... lLlbd7 looks
safer) 9 i.d3 (Hebert suggests 9 'iVe2,
with the point 9 . . . i.g7 10 eS ! , but
9 . . .i.xbS ! 1 0 'iVxbS+ lLlbd7 improves)
and now:
c 1 ) 9 . . . a6?! 1 0 lLlf3 ! (Larin-Kap­
ustin, Moscow 1 997 went 1 0 a4 ! ?
i.g7 1 1 lLlf3 0-0 1 2 0-0 i.g4 with easy
equality) 1 O...i.g4 ( l O . . .bS 1 1 eS b4
12 lLle2 lLlxdS 1 3 i.e4) 1 1 0-0 i.g7
( l l . ..lLlbd7 12 'iVa4 ! ? i.xf3 1 3 l:hf3
.Jtg7 1 4 eS !) 1 2 eS lLlfd7 1 3 e6 ! fxe6
14 dxe6 i.xe6 I S .l:tel with a terrific
attack.
c2) Better is 9 ...i.g7 1 O 'iVb3 ! ? bS !,
a sacrifice resembling other lines in
this book. Play might go 1 1 lLlxbS
.JtxbS I 2 �xbS+ lLlbd7 1 3 lLlf3 lLlxe4! ?
14 0-0 fS with chances for both sides,
and very hard to assess. Still, I feel this
should be fine for Black.
S i.g4 (D)
.•.
8 .. . lLlbd7? is a known error because
9 eS ! dxeS 10 fxeS lLlxeS 1 1 i.bS+! is
107
extremely strong; e.g., 1 1 .. .lLled7+ 12
'ifilf2 ! lLlg4+ 1 3 'ifilg3 +-.
9 h3
Or:
a) After 9 i.bS+ lLlbd7 1 0 0-0, I
like 1 O ...0-0-0! ?, partly because you
don' t often get to castle queenside in
the Modem Benoni ! For example, 1 1
'iVa4 ( 1 1 l:1el i.g7 1 2 eS dxeS 1 3 fxeS
lLlxeS 1 4 1heS 'iVxeS I S lLlxeS i.xdl
1 6 lLlxf7 i.c2 +) 1 1 . . . 'ifilb8 1 2 eS lLle8
=. Nevertheless, 1O ... a6 may not be so
bad, since 1 1 eS, as in A.Davie-Aitken,
British Ch (Hastings) 1 965, can be met
by l l .. .axbS ! 1 2 exf6 ( 1 2 lLlxbS 'ifild8)
12 . . .'iVxf6 with the idea 1 3 lLlxbS 'ifild8.
b) 9 i.d3 is a natural continuation:
9 . . .i.g7 1 0 0-0 (after 10 eS, 1 O. . .dxeS
1 1 fxeS i.xf3 1 2 �xf3 lLlbd7 ! ? looks
adequate; e.g., 1 3 0-0 lLlxeS 14 i.bS+
lLlfd7 I S 'iig3 0-0 1 6 i.f4 :fe8 1 7
�fel J:lad8 1 8 l:.ad l a6 1 9 d6 'iVf8 20
i.fl lLlc6 +) 1 0 ... 0-0 and now:
b l ) 1 1 eS lLle8 ! ? 1 2 l:te1 i.xf3 1 3
'iVxf3 dxeS 1 4 fxeS i.xeS I S i.f4
lLld7 ! 1 6 i:tad l ( 1 6 i.bs lLld6 1 7 i.. xd7
i.d4+) 1 6 . . .lLld6 ! .
 108 THE GAMBIT GUIDE TO THE MODERN BENONI
b2) 1 1 h3 i.xf3 1 2 'it'xf3 lLibd7 1 3
l:te1 ( 1 3 eS dxeS 1 4 fS e4 ! ) 1 3 . . . a6
( 1 3 ...lLie8 !?; 1 3 ... l:tab8 ! ?) 14 a4 lLie8
I S i.e3 ( IS i.d2 lLic7 1 6 as :tab8 1 7
l:ta2 bS ! 1 8 axb6 l:txb6 =) l s ...lLic7 1 6
l:tad 1 bS ! 1 7 axbS lLixbS 1 8 lLixbS
axbS 1 9 i.xbS .:tfb8 ! 20 'it'e2 liaS =.
Admittedly, Black is walking a tight­
rope in such lines, but that doesn't
mean that they're bad for him!
c) 9 i.e2 lLibd7 and now:
e l ) 10 h3 i.xf3 1 1 i.xf3 i.g7 1 2
0-0 0-0 (this closely resembles the
9 ...i.g4 variation of the Four Pawns
Attack) 1 3 l:te1 l:tfe8 14 lLibS ! ? lLib6
IS eS dxeS 1 6 d6 'it'd7 ! 1 7 lLic7 e4 1 8
i.e2 l:ted8 ( 1 8. . .l:tad8 1 9 i.bS) 1 9
lLixa8 lIxa8 with compensation; for
example: 20 i.e3 lLifdS 2 1 i.xcs i.xb2
22 :b1 lLic3 23 'it'c2 lLixb1 24 l:txb1
(24 'it'xb2 lLia4 =t) 24. . .i.g7 2S 'it'xe4
lLia4 =.
c2) 10 0-0 i.xf3 ( 1 O ... lLixe4 ! ? is a
little unlikely, but may be worth a
look: 1 1 lLixe4 { I I :tel lLixc3 1 2 bxc3
0-0-0 is unclear } 1 l . . .'it'xe4 1 2 :e1
i.e7 1 3 lLigS i.xe2 1 4 l:txe2 'it'd4+ I S
i.e3 'it'xd 1 + 1 6 .:txd 1 0-0) 1 1 i.xf3
0-0-0 1 2 l:te1 ( 1 2 'ii'a4 a6; 12 eS dxeS
1 3 lLibS a6 14 d6 'it'e6 =t, intending to
meet I S fxeS with l S ... axbS 16 exf6
lLieS) 1 2 ... a6 1 3 eS dxeS 1 4 fxeS lLixeS
IS 'iii'h 1 i.g7 1 6 i.f4 lLifd7 17 .:te l
l:the8 - unclear?!
d) 9 eS, as always, is a critical chal­
lenge: 9 ...dxeS 1 0 fxeS i.xf3 1 1 1i'xf3
( 1 1 i.bS+ lLibd7 1 2 1i'xf3 1i'xeS+ 1 3
'iii'f1 { 1 3 'iii'f2 'it'd4+ } and now either
1 3 ... 0-0-0! ? or 1 3 ... i.d6 1 4 i.h6 'iii'd8)
1 l . . .'it'xeS+ ( 1 1 . . .lLibd7 ! ?) 12 i.e2
i.d6 1 3 i.h6 ( 1 3 g3 0-0 14 .if4 'it'e7
I S .igS lLibd7 1 6 0-0 'iii'g7 1 7 l:tae 1
i.eS) 1 3 ... lLibd7 14 0-0-0 0-0-0 I S
lLibS .ib8 1 6 g3 l:the8 1 7 :the 1 1i'e4 =t.
This is a little scary ! Fortunately, Black
isn't the only one who is at risk here.
9 i.xf3 10 'ii'xf3 i.g7 (D)
.•.
w
We have a unique kind of Four
Pawns Attack that doesn't arise by any
normal KID move-order. I didn't like
this kind of set-up at all when I first
wrote about the Taimanov Attack. But
now that I see its similarities with the
...i.g4 line versus the Four Pawns At­
tack (A2 1 ), I am encouraged. In com­
parison with that line, Black's rooks
are quickly connected, and a timely
... lLie8 can be handy.
11 .id3
1 1 .ic4?! lLixe4 1 2 lLixe4 fS 1 3
i.d3 0-0 gives Black at least equality.
11 ... 0-0 12 0-0 lLibd7 13 .id2
The other bishop move also makes
sense: 1 3 .ie3 a6 1 4 a4 :tfc8 (after
1 4 ...:tab8, Black can be happy with I S
eS? dxeS 1 6 fS e4 1 7 i.xe4 lLixe4 1 8
1i'xe4 1i'xe4 1 9 lLixe4 .:tfe8 20 lLixcs
:txe3 2 1 lLixd7 :td8 + with the idea 22
 PAWN-STORM SYSTEMS
lbb6 iLd4; however, I S iLf2 ! is hard
to meet) I S lIfe l nab8 1 6 iLf2 lbe8
with a typical position from the ...iLg4
Four Pawns Attack. Black aims for his
customary . . .lbc7 and ... bS, even if the
opponent chooses to play aS. White is
probably somewhat better, but he has
some trouble finding a coherent plan.
13...a6
1 3 ...lbe8 14 ':fel lbc7 ( 1 4 . . . a6) I S
'it'h l �h4? ( I S . . .nab8 !) 1 6 :e2 nae8
17 iLel �d8 1 8 iLg3 ! gave White a
nice advantage in Panno-Aitken, Mu­
nich OL 1 9S8.
14 a4 nab8 15 iLe1
Or IS nfel lbe8 16 as ( 1 6 iLc4
lbc7) 1 6. . . lbc7 ( 1 6. . .bS ! ?) 17 eS dxeS
1 8 fS bS 1 9 axb6 ':xb6 = .
15...lbe8 16 iLg3 lbc7
Yet again Black poses the kind of
problem given in the note to White's
1 3th move. Play might continue 17 eS
( 1 7 .l:.ad 1 bS 1 8 axbs lbxbS =) 17 ...dxeS
1 8 fS c4 1 9 iLe4 and now 1 9 ...'iVd6 is
equal. Less clear but worth consider­
ing is 1 9 . . .bS ! ? 20 axbS lbxbS 2 1
z::.xa6 lbd4.
Naturally, the above analysis only
scratches the surface. It's hard for me
to believe that 7 . . .'iVe7 is fully equal
against perfect play, but it seems close,
and offers good practical chances.
Such an unknown line also has consid­
erable surprise value. The drawback is
that Black risks being overrun by a vi­
olent tactical onslaught, but I haven't
found anything so dire. I think that
7 . . .�e7 will probably appeal most to
those who fear the Taimanov Attack.
109
Of the f4 systems in this chapter,
the least effective is the Mikenas At­
tack, which allows for multiple solu­
tions and is not truly dangerous. Since
the Four Pawns Attack is fundamen­
tally a King's Indian Defence, I have
suggested two less analysed responses,
and I regard 9 i.e2 lbbd7 (Line A22)
as a particularly exciting and promis­
ing line.
In the important Taimanov System,
my proposed solutions obviously cover
a lot of new ground. Many of the criti­
cal continuations rest mainly upon
analysis, and the play tends to be too
chaotic to yield definitive conclusions.
Some players may be disturbed by
this, as well as by the undeniable dan­
ger to which Black is exposed. In re­
sponse, I should first point out that a
degree of risk goes with the territory,
i.e., when one plays the 'pure' Benoni
(without lbf3 in), one must be particu­
larly willing to enter into unruly com­
plications. Indeed, Benoni players
tend to revel in a bit of chaos. Remem­
ber that White is similarly subject to
danger and uncertainty. It is also im­
portant to realize that there is plenty of
room for creativity in this relatively
unexplored territory. One certainly
needn't follow my analysis inflexibly,
and I have also left numerous alterna­
tives for Black unanalysed, such as
ideas with ...lba6 and ...lbb4. My gen­
eral conclusion is that one can have a
lot of fun with the lines in this chapter
and that they will lead to positions rich
in possibilities.
6 Fia nchetto Systems with g3
1 d4 lLlf6 2 c4 cS 3 dS e6 4 lLlc3 exdS 5
cxdS d6 6 lLlf3
6 g3 g6 7 .i.g2 .i.g7 (D) normally
ends up transposing to our main line,
though there are a few independent
paths:
w b3) 9 ... l:te8 10 0-0 ( 1 0 a4 lLla6! 1 1
a) White can try 8 lLlh3, but if lLlf4
follows, it obstructs his pieces: 8 . . . 0-0
9 0-0 b6 (or 9 . . . a6 1 0 a4 .txh3 ! ? 1 1
.txh3 lLlbd7 1 2 .tf4 'fic7 1 3 :te l
l':.ae8 14 .i.g2 lLleS = Vadasz-Varga,
Budapest 1 998) 1 0 lLlf4 a6 1 1 a4 !!.e8
12 h3 l:la7 ! 1 3 e4 ':'ae7 = Moeller­
Baklund, Moscow 1 9S6.
b) The idea 8 e4 0-0 9 lLlge2 fails to
control the eS-square and forfeits the
idea of placing a knight on c4. Black
has several good answers, of which
lines 'b2' and 'b3' are best, in my
opinion:
b l ) 9 . . .lLla6 10 0-0 lLlc7 1 1 a4 a6
12 as l:te8 1 3 f3 lLlbS 14 .i.e3 lLld7 =
(Kapengut).
b2) 9 . . . a6 1 0 a4 lLlbd7 1 1 0-0 l:lb8
1 2 as bS ! ( 1 2 ...l:le8 =) 1 3 axb6 �xb6 !
(since lLlc4 isn't available) 14 h3 and
now 1 4 . . . l':.e8 = Novotelnov-Plater,
Moscow 1 947, or 1 4 ... lLle8 with the
idea . . .lLlc7-bS-d4 - Kapengut.
0-0 lLlb4 1 2 h3 b6 1 3 .tgS h6 14 .te3
.iL.a6 = Tanin-Aratovsky, COIT. 1 9S4)
1 O...bS ! 1 1 lLlxbS lLlxe4 12 .txe4 lhe4
1 3 lLlec3 :e8 1 4 .tf4 .i.f8 = Murei­
Quinteros, New York 1 983.
6 g6 7 g3 .tg7 8 .i.g2 0-0 (D)
..•
9 0-0
This is the basic position of the
Fianchetto System, a very solid ap­
proach which at first glance uses two
 FIANCHETTO SYSTEMS WITH g3
tempi to put the king' s bishop on a bad
diagonal (blocked by the dS-pawn),
and doesn't even begin to expand in
the centre. Yet this is a popular choice
of strong positional players, because it
covers e4, protects the king, and al­
lows a sort of Knight' s Tour by ttJd2-
c4. Often, when Black turns his atten­
tion to challenging that knight, White
can play moves like l:te 1 and .i.f4 fol­
lowed by e4-eS , which incidentally
can bring the g2-bishop to life.
White's strategy is primarily pre­
ventative: he advances few pawns and
denies Black an easy target. Key cen­
tral squares are to be guarded so as to
hinder any freeing moves, and even
manoeuvres such as ... ttJg4-eS are of­
ten discouraged by h3.
Since Black lacks space, his inabil­
ity to grapple directly with the enemy
can be irritating for him. But the very
nature of White's scheme also gives
Black more leeway, because his own
position is under no attack. White
tends to have his pieces on the first and
second ranks (e.g., rooks on e l and a i ,
bishops on c 1 and g2, knights on c3
and d2, queen on c2). This means that
Black can organize his troops into
whatever formation he pleases, since
he is under little obligation to defend
the usual sore spots on d6 and eS . And
fortunately for him, he still has one
plan that can't be prevented forever,
i.e., the customary advance of his
queenside majority. To the extent that
White must divert his forces to keep
the queenside pawns in check, Black
is given greater freedom to manoeuvre
in the centre.
111
There are fewer issues of move­
order in this chapter than in any other.
Instead of 9 0-0, the only non-trans­
positional sequence that I'm aware of
is the immediate 9 ttJd2 a6 10 a4 ttJbd7
1 1 ttJc4 ( 1 1 0-0 :e8 transposes to
Line A) 1 l .. .ttJb6 1 2 ttJa3 .i.d7, which
can lead to:
a) 1 3 .i.d2 ttJxa4 ! 14 ttJxa4 bS I S
ttJc3 ( I S ttJxcS dxcS 1 6 d6 { 1 6 0-0
l:te8 } 1 6 ....:c8; IS .i.f4 bxa4 16 .i.xd6
l:te8) l S . . .b4 16 ttJc4 bxc3 1 7 bxc3
'Vitie7 1 8 ttJb6 l:ta7 1 9 ttJxd7 ttJxd7 20
0-0 :b8 = Krasenkov-Tolnai, Buda­
pest 1 989.
b) After 1 3 0-0, 1 3 . . .l:.e8 transposes
to Line A; if Black wants to avoid
those lines, he can also try 1 3 . . .:b8 1 4
a s ttJc8 I S ttJc4 and, for example,
I S ... ttJe8 ( 1 S .. Jie8 again transposes to
Line A) 1 6 .i.f4 gS ! ( 1 6 ... fS 1 7 'Vitid2
.i.bS 1 8 ttJxbS axbS 1 9 ttJa3 ttJc7 =) 1 7
.i.d2 fS 1 8 lic 1 h6 1 9 h4 .i.f6 2 0 hxgS
hxgS 2 1 f4 g4 22 b3 lie7 =t.
c) 13 as ttJc8 1 4 ttJc4 ttJg4 IS .i.f4
.i.bS ! ? 1 6 'ttVb3 .i.xc4 1 7 lixc4 ttJeS
( 1 7 . . .bS ! ? 1 8 axb6 ttJxb6) 1 8 'ttVa2 bS
19 axb6 ttJxb6 20 0-0 lie7 (20 .. Jle8)
2 1 b3 1:tfb8 ! ? (2 l ...ttJd3 ! ? 22 exd3
.txc3 23 l:tac 1 .i.b4 is unclear) 22
'ttVc2, Krasenkov-Petran, Balatonber­
eny 1 988, and now simply 22 . . .aS with
the idea of ... c4 keeps Black active, al­
though his earlier options achieve
equality more easily.
All these examples bear close re­
semblance to the main lines below.
After the text-move (9 0-0), play di­
vides into:
A: 9...a6 10 a4 ttJbd7 1 12
B: Lines with ... .:.e8 1 24
112 THE GAMBIT GUIDE TO THE MODERN BENONI
The first approach is a positional
one, calmly preparing the eventual
queenside attack, although when the
goals of the two sides conflict, the
usual Benoni fireworks can erupt. The
second strategy is impatient and tacti­
cal. Black will enforce moves like
...lbe4 before White has had a chance
to organize his preventative forma­
tion.
A)
9 a6 10 a4 lbbd7 (D)
•.•
This is the most-played line of the
g3 fianchetto system. White has these
main moves:
AI: 11 e4 1 13
A2: 11 i.f4 1 14
A3: 11 lbd2 1 15
The last is the most important. But
many top players choose 1 1 i.f4 to
avoid the complexity and depth of the­
ory on 1 1 lbd2, so that will also be
studied in detail.
Note that 1 1 e4 cannot arise from the
9 ... a6 10 a4 1Ie8 move-order, but in that
case 1 1 lbd2 sometimes transposes, as
explained in the relevant section (Line
B2).
Also played are:
a) 1 1 a5 :e8 ( l l ...b5 12 axb6 lbxb6
1 3 e4 11e8 1 4 .:%.el , Pons-Gil Reguera,
Madrid 1 992, and now 1 4 ... h6 ! ? 1 5
'iWc2 l:ta7 ! ? 1 6 l:td l l:tae7 1 7 lbd2 lbg4
is double-edged) 1 2 i.f4 'iWc7 1 3 lba4
b5 ! 14 axb6 lbxb6 1 5 lbxb6 "iixb6 1 6
lbd2 was given as ± by Marin, but I
would be happy with 1 6..."iix b2! 1 7
lbc4! ? ( 17 i.xd6 i.g4 ! ; 1 7 .:%. b l 'iWa3 ! ?
1 8 .i.xd6 i.g4 1 9 f3 "iie3+) 1 7 ..."iixe2
1 8 lbxd6 "iix dl 1 9 1Ifxdl .:%.d8 = or
1 9 ....:%.e2 =.
b) 1 1 h3 l:te8 (perhaps the simplest
move; both 1 l . ..l:tb8 and l l ...h6 also
equalize, according to theory) :
bI) 1 2 i.f4 lbh5 1 3 .i.g5 ( 1 3 .i.xd6
"iib6) 1 3 ...f6 14 .i.d2 f5 1 5 i.g5 .i.f6
( 1 5 ... 'iWb6 1 6 "iid2 lbdf6 1 7 a5 "iic7 =)
16 'iWd2 lbe5 = .
b2) 12 "iic2 b6 13 l:%.bl 'iWc7 14 i.f4
lbh5 1 5 .i.d2 lbe5 = 1 6 g4 ! ? lbxf3+ 1 7
i.xf3 lbf6 1 8 b4 lbd7 1 9 bxc5 lbxc5
is equal, Koskinen-Sher, Copenhagen
1 996.
b3) 1 2 1Iel lbe4 (this resembles
the .. .lIe8 systems covered in Line B ,
but here h 3 i s of little use t o White) 1 3
lbxe4 l:txe4 1 4 "iic2 "iie7 ! ? ( 14. . .lIe8
= ) 1 5 .i.d2 lbf6 16 lbh4 i.d7 ! 17 i.c3
1Ie8 1 8 .i.xe4 lbxe4 19 .i.xg7 �xg7
20 1la3 (20 Wg2 lbg5 21 g4 h5 22 f3
lbh7 = ), Razuvaev-Tal, USSR Ch
(Moscow) 1 983, and now 20...i.xh3 =
is easiest, with two pawns for the ex­
change.
c) 1 1 1Ib 1 .:%.e8 1 2 b4 ( 1 2 .i.d2 lbb6 !
1 3 e4 lbc4 14 .i.c l b5 1 5 axb5 axb5 1 6
 FIANCHETTO SYSTEMS WITH g3
b3 liJa3 ! 17 .i.xa3 b4 + Lehman-Har­
toch, Switzerland 1 97 1 ) 1 2 . . .'ii'c7 1 3
.i.d2 liJb6 1 4 liJg5 .i.f5 1 5 bxc5 'ili'xc5
1 6 ':'c 1 liJc4 ( 'unclear' ECO) 1 7 h3
- 'ili'c7 1 6 a5 .i.d7 1 7 liJa4, Brennink­
�6 1 8 .i.e l liJe3 ! + P.Nikolic-Vel­
imirovic, Yugoslav Ch 1 983.
d) 1 1 'ii'c2 :b8 ( l l . . .:e8 12 e4
filc7 transposes to note 'b' to White's
l i th move in Line B2) 12 a5 b5 1 3
axb6 liJxb6 1 4 liJd2 liJfd7 ! ? 1 5 liJb3
liJc4 1 6 liJd2 liJcb6 = Rohde-D.Gure­
vich, New York 1 985.
e) 1 1 :te l 'ili'c7 ( l l ...:tb8 is consid­
ered completely equal as well) 12 liJd2
:b8 1 3 a5 b5 14 axb6 liJxb6 1 5 ':'a2
liJfd7 1 6 'ii'c2 c4! 1 7 liJde4 h6 1 8 .i.f4
liJe5 = Alburt-Hjartarson, Philadel­
phia 1 986.
At)
11 e4 :e8
Also played is 1 l ...:b8 1 U te 1 b5.
12 ':'el c4! (D)
1 2 ...liJg4 is an extremely common
move here, but I prefer this aggressive
approach.
13 .i.fl
113
Other moves are answered tacti­
cally:
a) 1 3 h3 liJc5 1 4 liJd2 liJd3 15 ':'e2
meijer-de Firmian, Lugano 1 989, and
now I like just 17 ...liJxc l ! 1 8 'ili'xc l
liJxd5 +.
b) 1 3 .i.f4 liJc5 14 e5 liJg4! 1 5 e6?
fxe6 1 6 dxe6 liJd3 + Ligterink-Van
Blitterswijk, Leeuwarden 2000.
c) 1 3 'ili'e2 liJc5 ! intending 14 'ili'xc4
b5 1 5 axb5 axb5 1 6 1Wxc5 dxc5 1 7
':'xa8 b4 + Ligterink.
13...'iIi'c7 14 'ili'd4
1 4 1We2 liJc5 1 5 'ii'xc4 transposes to
the main line.
14...liJc5 15 'ili'xc4 .i. g4 16 .i.g2
Black also achieves good play fol­
lowing 16 liJd2 liJfd7:
a) 17 e5? :txe5 1 8 .tIxe5 liJxe5 1 9
'ili'f4 g5 ! 20 'ili'b4 (20 'ili'xg5?? liJb3 ! 2 1
liJxb3 liJf3+) 20. . ..i.f5 + threatening
. .. liJd3, Piket-Winants, Wijk aan Zee
1 987.
b) Black has compensation after 17
1Wb4 f5 18 f3 and now either 18 ... .i.h5
1 9 Wg2 liJe5 20 :te3 a5 ! ? 2 1 'ili'b5
:f8 ! 22 exf5 gxf5 = or 1 8 ... .i.h3 ! ? 1 9
.tIe3 .i.xfl 20 �xfl liJe5 =.
c) Similarly, 17 "ii'a2 f5 18 h3 .i.h5
intending moves like ...1;!ac8 and . . .g5
is quite satisfactory.
16 ...liJfd7 17 .i.f4 liJb6 18 'ili'fl
1 8 'ili'b4 liJd3 1 9 'ii'xd6 1Wxd6 20
.i.xd6 liJc4! + Velickovic.
After the text-move ( 1 8 'ili'fl), Els­
ness-Emms, Gausdal 1 995 continued
1 8 ... .i.xc3 1 9 bxc3 liJxe4 20 h3 .i.d7
2 1 'ili'd3 liJc5 +. The line beginning
with 1 2 . . .c4 is a good illustration of
Benoni dynamism.
114 THE GAMBIT GUIDE TO THE MODERN BENONI
A2)
11 i.f4 'ike7 (D)
In this kind of position, I prefer the
queen on c7, supporting queenside
counterplay and, after ...l:tb8 and ...bS,
covering a7.
1 l . ..'ike7 ! ? is also played, when a
few ideas are shown by 1 2 h3 ! ( 1 2
'ikd2 lZ)g4! 1 3 h3 lZ)geS =) 1 2. . .l:te8 1 3
e4 ( 1 3 'ikd2 l:tb8) 1 3. .. b6! ? ( 1 3 ... l:tb8
14 %leI lZ)hS I S i.gS 'ikf8 may be
slightly more accurate) 1 4 %leI lZ)hS
I S i.gS "ii'f8. Probably White has an
edge here, but it's not so easy to dem­
onstrate.
w = ) 1 4 . . . h6 I S i.e3 lhe3 ! ? 1 6 fxe3
12 h3
As usual, there are quite a few im­
portant alternatives:
a) 1 2 e4 l:te8 13 'ii'c2 ( 1 3 l:te1 lZ)g4
1 4 'ikc2 lZ)geS = Ruck-Stevie, Pula
1 999) 13 ...lZ)g4!? Schneider; this looks
equal. Black could also play 1 3 ...l:tb8
14 aS lZ)hS I S i.e3 bS 1 6 axb6 lZ)xb6
1 7 h3, Polugaevsky-Vaiser, Sochi
1 98 1 , when 1 7 ... fS ! ? is intriguing.
b) 12 'ikd2 .:te8 1 3 h3 ( 1 3 .:tfe l has
been successfully answered by 1 3 ...c4,
1 3 ...lZ)b6 and 1 3 ... lZ)hS) 1 3 ... c4 (ECO
gives 1 3 .. J�b8) 1 4 i.h6 i.h8 I S 'ikc2
lZ)cS? ! ( 1 S .. Jlb8 !) 1 6 lZ)d2 lZ)cd7 1 7
e4 ;t Conquest-Hjartarson, Hastings
1 985/6.
c) 1 2 .:te l has several adequate re­
sponses:
c l ) 12 ...'ii'b8 ! ? 1 3 h3?! ( 1 3 as bS 14
axb6 lZ)xb6 is double-edged) 13 ...bS
1 4 axbS axbS + Freeman-Giulian,
British Ch (Ayr) 1 978.
c2) 1 2 . . .l:le8 13 b4 ( 1 3 h3 l:tb8 14
�h2, Scherbakov-Kovacevie, Belgrade
2000, and now Black should play con­
sistently with 14 ...bS I S axbS axbS 1 6
b4 'ikb6 ! ; e.g., 1 7 bxcS lZ)xcs or 1 7
i.e3 l:txe3 ! , which looks very good for
Black) 1 3 ... lZ)hS 1 4 i.gS ( 1 4 i.d2 b6
cxb4 1 7 lZ)e4 'ikb6 1 8 lZ)d4 as gives
Black compensation: 1 9 'ikc2 fS 20
lZ)d2 (20 lZ)f2 lZ)cs 2 1 g4 ! ? lZ)f4 ! 22
gxfS b3 ! 23 'ikc4 lZ)xg2 24 �xg2
i.xfS ! with more than enough for the
exchange, P.Nikolic-Hjartarson, Lin­
ares 1 988) 20 ...i.xd4 2 1 exd4 'iWxd4+
22 e3 ! 'it'xe3+ 23 �h 1 , Krause-Lin­
demann, COIT. 1 994, and now Krause
gives 23 . . . lZ)eS ! 24 'ikc7 ! 'iWxd2 2S
'iVxd6 lZ)d7 26 'iVe6+ (not, however, 26
'iVxg6+? lZ)g7) 26 ... �f8 27 'iVd6+,
with a draw.
c3) 1 2 . . . lZ)hS 1 3 i.gS l:te8 ( 1 3 . . .h6
has also equalized) 1 4 'iid2 lZ)hf6
(White's pieces are now a little mis­
placed) I S i.h6 i.h8 1 6 h3 lZ)e4 1 7
lZ)xe4 %lxe4 1 8 b4 'iVd8 ! 1 9 l:tb 1 as !
20 bxaS %lxaS with a slight advantage
for Black, Drasko-Ve1imirovic, Subo­
tica 2000.
12 l:te8 (D)
•..
 FIANCHE1TO SYSTEMS WITH g3 115

Or 12 ...:b8 13 l:k l h6 1 4 e4 :e8 1 9 ...�d8 20 g4 tDf6 2 1 tDd4 hS 22 gS

I S l:Iel gS 16 i..d2 bS = Blees-Dam­ tLlh7 23 tDf3 i..fS =.
bacher, Rotterdam 1 999.
A3)
11 tDd2
This is the critical variation of the
Fianchetto System, so I will offer two
lines to choose from:
A31: 11 ..Jle8 l I S
A32: l1 ...tDhS 1 23

A3 1 )
11 ...l:te8 (D)

W
13 l:tel
1 3 �c2 l:tb8 (a good alternative is
13 . . . tDh5 14 i..d2 tDeS = I S g4 ! ?
tDxf3+ 1 6 i..xf3 tDf6 1 7 b3 hS ! 1 8 gS
tLlh7 and Black is slightly better,
Chiburdanidze-Wang Yu, Shenyang
FIDE worn World Cup 2000) 14 as bS
I S axb6 tDxb6 1 6 e4 tLlhS 17 i.. c l
(other retreats allow ...tLlc4) 1 7 ...tDf6
( 1 7 . . . fS ! ? might be better) 1 8 'ith2
tLlc4, perhaps lightly 1, Korchnoi­ Now there are two very important
Haapasal0, Copenhagen 1 996. answers, leading to a final split:
13...tDhS A311: 12 tDc4 1 1 7
1 3 . . .l:.b8 14 e4 tLlhS intending . . .bS A312: 12 h3 1 18
has also equalized in several games.
After the text-move ( l 3 . . . tLlhS), The numerous alternatives tend to
Razuvaev-Benjamin, Paris Ch 1 989 be less challenging, but they contain
continued 14 i..d2 c4 ( l 4 . . .l:tb8 IS as some essential ideas and are well
bS 16 axb6 tDxb6 intending ...tDc4 is worth playing through:
also reasonable) IS b4 (else ... tDcS-b3) a) 12 �b3 :b8 ( l 2 ...tDeS 1 3 f4 ! ?,
I S ... cxb3 1 6 'i\Vxb3 tDcs = 1 7 �b4. contrary to ECO, is just fine for Black
Then I think that 1 7 ...i..d7 is the easi­ after 1 3 ...tDeg4 ! 14 tDc4 tLlhS I S tDe4
est way to play; for exm;nple, 1 8 l:tac 1 fS !) 1 3 ttJc4 ttJeS 14 tLlb6 tDfd7 (or
I1ac8 1 9 as ( 1 9 g4 as ! ? 20 'iVc4 tDf6 =) 14 ...tLled7 I S tDxc8 'ilYxc8 1 6 as bS 17
116 THE GAMBIT GUIDE
axb6 l:lxb6 = Johansen-Rogers, Syd­
ney 199 1 ) I s lDxc8 �xc8 16 ..th3 ( 1 6
h 3 bS ! ? 1 7 axbS axbS 1 8 l:!.aS b4 19
lDe4 "fIc7 is equal, Sakaev-Anastas­
ian, Frunze 1 989) 1 6 . . . fS, Ionescu­
Stoica, Romanian Ch 1 98 1 . Now Stoica
gives 1 7 f4 lDf7 1 8 e4 (else we have
the familiar advantage of Black's e4
outpost and White's bad c l -bishop)
1 8 ...c4 19 �c2 lDcS ! 20 exfS lDd3 21
fxg6 'ilYxh3 22 gxf7+ �xf7 as 'un­
clear' but this is =+= or worse, since
White's king is too exposed.
b) 1 2 as bS ( 1 2 ... l:!.b8 1 3 lDc4 lDeS
14 lDb6 lDed7 = has also occurred) 1 3
axb6 lDxb6 1 4 lDb3 (heading for as
and c6) 1 4... ..td7 ( 14 ...lDc4 = is often
played, but lDaS-c6 turns out not to be
a problem) I S lDaS fic7 (or I S ... ..tbS
16 �c2 lDfd7 1 7 l:!.el lDeS = Ja.Bol­
bochan-Maderna, Argentine Ch (Bue­
nos Aires) 1 9S3) 16 l:!.el lDg4 17 h3
lDeS 1 8 f4? ! lDec4 19 lDxc4 lDxc4 20
'ilYd3 lDe3 ! 2 1 ..te4 (21 ..txe3 nxe3)
2 1 . ..fS 22 ..txe3 fxe4 23 'ikc4 'ilYb7 24
l:ta3 �b4! 2S �a2 ..td4! 26 ..txd4
cxd4 with a winning position for Black,
Klee-Lau, Dresden 1 996.
c) 1 2 l:ta2 l:!.b8 1 3 as bS 14 axb6
lDxb6 I S b3 lDfd7 ( 1 S ... lDg4 16 ..tb2
ilie7 1 7 e4 lDeS 1 8 'ikc2 c4 ! '+ Kekki­
Kuczynski, Pohja 1 985; I s ... lDhS 16
..tb2 fS 1 7 e3 lDf6 1 8 ..tal �e7 1 9 l:!.el
lDg4 ! , at least =, Dautov-Oll, Kiev
1 984) 1 6 'iWc2 fS 1 7 ..tb2 lDf6 1 8 l:!.d l
fie7 = Roos-Moiseev, Berlin 1 994.
d) 1 2 ne l (a move seldom seen at
this juncture, since on d2 the knight no
longer supports e4-eS) 1 2 .. .l::t b 8 1 3
lDc4 ( 1 3 as bS 1 4 axb6 lDxb6 is a type
of position we have seen many times;
TO THE MODERN BENONI
e.g., IS h3 lDfd7 16 �h2 fie7 1 7 lDde4
h6 1 8 ..tf4 lDc4! '+ Berezin-Brodsky,
Swidnica 1 999) 1 3 . . .lDb6 ( 1 3 ... lDeS
14 lDxeS l:xeS =) 14 lDa3 ..td7 (like­
wise, a standard manoeuvre) I S as
lDc8 1 6 lDc4 ..tbS 1 7 fib3 ..txc4 1 8
fixc4 lDd7 = Kruck-Zawadzki, Lito­
hoto 1 999.
e) 12 'iVc2 l:!.b8 13 lDc4 ( 1 3 as bS
14 axb6 lDxb6 IS e4 lDg4 = Krasen­
kow-Gritsak, Suwalki 1 999) 1 3 ...lDeS
1 4 lDxeS l:!.xeS I S as bS 16 axb6 1hb6
17 e4 ..td7 1 8 ..td2 l:!.e7 1 9 h3 fib8 =
20 l:.a2 lDe8 2 1 l:!.fal lDc7, intending
. . .fS or . . .lDbS-d4, Liberzon-Stein,
USSR Ch (Tbilisi) 1 966/7.
f) 1 2 e4 (D).
B
1 2... l:tb8 (or 1 2 ...fic7 1 3 lDc4 lDb6
14 lDa3 .i.d7 I S aS lDc8 1 6 lDc4 .i.bS
=) 1 3 lDc4 lDb6 ( 1 3 . . .lDeS 14 lDxeS
':'xeS I S as? ! bS 16 axb6 l:!.xb6 '+ Con­
quest-Sigurjonsson, Brighton 1 98 1 )
14 lDa3 lDa8 ! ? (an unusual idea, espe­
cially when White hasn't played as,
allowing ... lDc7-bS; 14. . ...td7 would
be normal) I S l:!.el lDc7 1 6 ..tfl bS ! ?
1 7 axbS axbS 1 8 lDaxbS lDxbS 1 9
 FIANCHETTO SYSTEMS WITH g3
.ltxb5 .ltd7 2 0 .ltxd7 �xd7 2 1 ..t>g2
tDg4 is unclear, Quinteros-Estevez,
Torremolinos 1 974. This reminds me
of a Benko Gambit, in spite of the dif­
ferent pawn-structure.
g) 1 2 .:lbl lIb8 1 3 b4 b5 ( 1 3 ...cxb4
14 lIxb4 tDc5 is also equal, as in
Soppe-Rosito, Saenz Pena 1 996 and
Chiburdanidze-Uskova, Istanbul worn
OL 2000) 14 axb5 axb5 = 15 bxc5? b4
16 tDce4 tDxc5 1 7 tDxc5 dxc5 1 8 e4
( 1 8 tDc4 .ltf5 1 9 .:lb3 .lte4) 1 8 . . ..lta6
19 'Wib3 ( 1 9 .:leI i.d3) 19 ...c4! 20 tDxc4
tDxe4 2 1 lie 1 tDc5 0- 1 Vaganian­
Tseshkovsky, USSR Ch (Leningrad)
1 974. This game illustrates how dev­
astating Black's piece activity can be.
A3 1 1 )
12 tDc4 tDb6
Theory approves 1 2 . . .tDe5 as well,
but the theory on it could fill a small
book. In my opinion, 12 . . .tDb6 is fully
satisfactory and provides plenty of
winning chances.
13 tDa3
Or:
a) 1 3 tDxb6 'Wixb6 14 as 'WIc7 is
harmless, as usual: 15 'Wic2 ( 1 5 e4 .ltd7
1 6 f4, Litwin-Ostrowski, Polanica
Zdroj 1 994, and now either 16 . . . tDg4
1 7 ..t>h l fS or 1 6. . .b5 is good) I S . . .lIb8
16 e4 tDd7 ( 1 6 . . . bS) 17 tDdl ! ? b5 1 8
axb6 tDxb6 1 9 tDe3, Krasenkov-Sher,
Moscow 1 990, 1 9. . ..ltd7 ! 20 .ltd2 .ltbS
== Schneider.
b) 1 3 tDe3 tDg4 1 4 tDxg4 .ltxg4 1 S
h 3 .ltd7 ( l 5 . . ..ltxc3 ! ?) 1 6 'Wid3 lIb8
with equality, Stahlberg-Kluger, Bu­
charest 1 9S4. .
13 i.d7 (D)
•••
117
This idea applies to a variety of
lines in more than one chapter. Black
both develops and frees c8 as a retreat
for the knight should White play aS. In
that case, the knight can go from c8
via a7 to b5, or it can spring back into
play following ...b5 and . . . tDxb6 (after
White's axb6). In the absence of as,
Black can play .. JIb8 to prepare, e.g.,
... tDa8-c7 and ... bS ; or he has interest­
ing tactical ideas based upon . . .tDxa4
and . . . bS, which are not so easy to pre­
vent.
14 .ltd2
Best, according to both ECO and
Schneider. The idea is to get b3 in at
the right moment. At any rate, nothing
else should scare Black:
a) 1 4 .ltf4?! tDh5 ! 1 5 .ltd2 (cer­
tainly not IS .ltxd6?? .ltxc3 ! 16 bxc3
.ltxa4) 1 5 . . .tDxa4 ! 1 6 tDxa4 bS. This is
a standard sacrifice, as we will see
from the many examples that follow.
b) 14 a5 tDc8 I S tDc4 .ltb5 16 �b3
i.xc4 ( 1 6 . . .�d7 ! ?) 1 7 'ii'xc4 b5 1 8
axb6 tDxb6 (even 1 8 . . .'ii'xb6 followed
by . . .tDd7 seems adequate) 1 9 �h4
( 1 9 'Wid3 'WIe7) 1 9 . . .tDfd7 20 .ltg5 (20
118 THE GAMBIT GUIDE TO THE MODERN BENONI
�xd8 l:texd8 2 1 lta2 ltdb8 22 ltd 1 as
23 e4 a4 = Minev-Hulak, Pula 1 975)
20 .. :i�b8 ! = Piankov-Psakhis, Irkutsk
1 977.
c) 14 h3 ltJxa4 ! ? (or 14 . . .ltb8) I S
ltJxa4 b S 1 6 ltJc3 b4 1 7 .i.d2 bxc3 1 8
..txc3 ltJe4 1 9 .i.xg7 �xg7 20 �c2 fS
2 1 ltJc4 .i.bS 22 b3 'it'f6 = Akopian­
TelCic, Belgrade 1 988.
d) 14 'iVc2 'ilVc7 I S h3 (1S .i.d2 .:tab8
1 6 h3 ltJxa4 ! 1 7 ltJxa4 bS is equal,
Spassov-Barlov, Vmjacka Banja 1 984)
I S . . . ltJxa4 ! (again !) 1 6 ltJxa4 bS 1 7
ltJc3 ( 1 7 ltJxcS �xcS 1 8 'iYxcS dxcS is
unclear or perhaps :t, because White's
pieces are awkwardly placed) 17 ... b4,
Loviscek-Bednarich, Nova Gorica
2000, and now White should bail out
with 1 8 ltJabi bxc3 1 9 ltJxc3 .i.bS =.
14 ...ltJxa4!
Why not? It seems there is hardly a
position in which this doesn't work.
15 ltJxa4 b5 16 ltJc3 b4 17 ltJc4
bxc3 18 .i.xc3 .i.b5 19 b3 (D)
B �xf6, but Black still has the advantage
19 ..ltJe4
. in earlier lines, and not yet committing
Played to improve upon 1 9 ...ltJg4
20 .i.xg7 r3Jxg7 2 1 e4 ltJeS 22 ltJxeS
l:.xeS 23 lte 1 , Lalic-Sax, Sarajevo
1 985, given as ;!; in several sources, al­
though this doesn' t seem obvious after
something like 23 . . .l:te7 ! ? 24 f4 (the
idea in that game) 24 ...f6 2S .th3 �6
26 �h 1 l::tae8 27 .i.e6 .i.d7 28 .i.xd7
':xd7.
Aside from all this, Fritz suggests
19 . . ..:ta7 ! ?, and indeed, it's hard to
believe that Black will be worse after
...ltae7. In general, Black has no real
problems in this line.
20 .i.xg7 �xg7 21 �c2 f5 22
�b2+?!
Remarkably, White gets in trouble
after this natural move. Uninspired but
better is 22 e3.
22 JlVf6!
••
Now:
a) 23 �c2? was played in D.Bekker
Jensen-Bjarnehag, Copenhagen 1 998,
where Black chose 23 . . .ltab8?, miss­
ing the opportunity for 23 ... .i.xc4 ! 24
bxc4 (24 .i.xe4 .i.xe2 ! ) 24 . . . ltJxf2 !
and Black is winning due to 2S �xf2
'iYd4+.
b) White should prefer 23 'iYxf6+
owing to the threat of . . . .i.xc4 fol­
lowed by ... ltJd2.
I believe that the line with 12 . . .ltJb6
is fully sound, and much easier to
learn than 12 ... ltJeS .
A3 12)
12 h3
This is a subtle move, taking away
the ... ltJg4-eS option we saw so often
to ltJc4.
12 l:.b8 (D)
.••
 FIANCHETTO SYSTEMS WITH g3
Black has had some success with
12 .. .lbhS, but I'll stick with the scheme
we used in the last section.
13 lDc4
This is most often played, although
1 3 as (line 'c') in particular has pro­
duced some very interesting play:
a) 1 3 lIbl lDhS (the idea is to
threaten ...bS without the possibility of
b4 in reply) 1 4 lDde4 ! ? ( 1 4 lDc4 lDeS
l S lDa3 fS 1 6 e4 bS ! 1 7 axbS axbS 1 8
b4 lDf7 = 10hannessen-Tisdall, Gaus­
dal l 99S; 1 4 lDce4 lDdf6 =; 14 'iWc2 bS
IS axbS axbS 16 b4 fic7 ! and the c3-
knight is a liability) 14 . . . lDdf6 IS g4
lDxe4 1 6 lDxe4 fS ! 1 7 lDxcS ? ! ( 1 7
CLlxd6 i s critical, but I think that Black
stands well; e.g., 17 ...'iIi'xd6 1 8 gxhS
gxhS 1 9 'iVc2 f4 ! 20 %1d l ? ! lIf8 { in­
tending . . ..tfS } 2 1 l:!.al .td4 ! with the
idea 22 'iVb3? f3 ! 23 exf3 .txh3 ! =+=)
17 . . . fxg4 1 8 lDe6 .txe6 1 9 dxe6 gxh3
20 .txh3 'ili'h4! =+= Lalic-Kovacevic,
Ossiach 1 984.
b) 1 3 lDde4 lDxe4 14 CLlxe4 CLlf6
( l4. . .lDeS ! ? , Schneider) I S .tgS .tfS
1 6 lDxf6+ .txf6 1 7 .txf6 li'xf6 1 8
119
'ii'd2 bS 1 9 axbS lIxbS 20 'ili'c3 112_1/2
Kuzmin-Grigorian, Minsk 1 976,
c) 1 3 as b5 14 axb6 lDxb6 (D) and
now:
c l ) I S CLlb3 lDc4 1 6 %1a4 'ili'b6 1 7
lhc4 'ili'xb3 = D e Briey-Alvis, Leon
1 996.
c2) IS lIa2 'ili'c7 16 b3 lDfd7 1 7
.tb2 fS 1 8 l:!.e l lIf8 ( 1 8 . . .SLb7 ! with
the idea .. ,lDf6) 19 f4 lDf6 20 �h2
c4 ! ? 2 1 b4 .tb7 22 lIaS 1He8 is equal.
Then, after 23 e3?, in Komljenovic­
Spraggett, Seville 1 994 Black missed
23 ...lDfxdS ! =+=; e.g., 24 SLxdS+ .txdS
2S lDxdS lDxd5 26 SLxg7 lDxb4.
c3) 1 5 'ili'c2 CLlh5 ( I S . . .'iVc7 is a
solid alternative) 1 6 'iVd3 ( 1 6 lIa2 fS
= ) 1 6 . . . .tfS ! 1 7 e4 .tc8 1 8 �h2 f5 1 9
f4 lDf6 = Skuja-Tal, Latvian C h (Riga)
1 9S8.
c4) I S e4 CLlfd7 1 6 f4 ( 1 6 l:!.e l lDeS
1 7 SLn c4 = Kosnar-Borkovec, Czech
Cht 1 99617) 1 6 ... c4! ? ( 1 6...fS 17 lIel
.td4+ 1 8 �h2 lDf6 = RO.Hernandez­
Holm, Siegen OL 1 970) 1 7 eS ? ! dxeS
1 8 fS? gxf5 1 9 l:!.xfS CLlcS =+= Cuevas
Rodriguez-Z.Franco, Cordoba 1 994.
120 THE GAMBIT GUIDE TO THE MODERN BENONI
13 ttJb6
..•
I ' m choosing this move to be con­
sistent with the 1 2 ttJc4 ttJb6 of Line
A3 1 1 . Most of the same themes cer­
tainly apply.
14 ttJa3
Or:
a) 1 4 ttJe3 .i.d7 I S i..d2 ttJhS 1 6 as
ttJc8 (or 16 ...ttJa8 1 7 c;t>h2 fS 1 8 f4 .i.bS
= Spasov-Rajkovic, Trstenik 1 978) 1 7
ttJc4 fS 1 8 l:.el ttJf6 1 9 �c2 'fic7 20
l:.abl and now, instead of 20 . . ..i.bS 2 1
ttJa3 .td7 = Kabisch-Kovacevic, Lu­
gano 1 989, 20 . . .bS ! 2 1 axb6 ttJxb6
gives Black the initiative.
b) 14 ttJxb6 �xb6 I S as �c7 is, as
always, innocuous: 1 6 e4 bS 17 axb6
lhb6 1 8 f4 l:.b4 = L.Schmidt-Gonda,
Tokyo 1 989.
14 i..d7 (D)
...
I trust this more than 14 ....i.fS ! ? I S
g4 i..d7 1 6 .tf4 hS ! 1 7 g S ( 17 .i.xd6
.txa4) 17 . . .ttJh7 1 8 h4 i.. xc3 1 9 bxc3
.txa4, which Van der Sterren calls
'unclear' .
w IS ... ttJc8 1 6 ttJc4 .tbS ( 1 6 . . .�c7 is the
We have reached a critical position
for this variation.
15 e4
This is now played almost exclu­
sively. Of the alternatives here, IS as
(line 'c') used to be popular:
a) I S 'ii'c2 is now met by the famil­
iar tactic I S . . .ttJxa4 ! ( 1 S . . .ttJc8 is also
playable) 1 6 ttJxa4 bS 17 ttJc3 ( 1 7
ttJxcS dxcS 1 8 e4 b4 1 9 ttJc4 b3 ! 20
�d3 l:.b4 2 1 i..d2 l:.xc4 22 'i¥xc4 i..bS
is 'unclear' - Magerramov) 17 . . .b4 1 8
ttJc4 ( 1 8 .i.d2 bxc3 1 9 i..xc3 was given
as equal by Magerramov; Black then
has 1 9 . . . i..fS !) 1 8 . . . bxc3 1 9 bxc3 ( 1 9
ttJxd6 cxb2 20 .i.xb2 1hb2 ! 2 1 �xb2
ttJe4 'with an attack' - Magerramov)
1 9 ... .tbS 20 ttJe3 ttJd7 2 1 l:tbl c4 =
Godzhaev-Magerramov, USSR 1 987.
b) IS .tf4?! ttJhS ! 16 .i.xd6 i..xc3 !
17 .txb8 ( 1 7 bxc3?? .txa4) 17 ....txb2
1 8 i..a7 ( 1 8 g4 �xb8 1 9 gxhS ttJxa4
20 l:.a2 ttJc3 2 1 �b3 ttJxe2+ 22 c;t>h l
i.. g7 + Hort-Marovic, Banja Luka
1 976) 1 8 . . .ttJxa4 ! T V.Sokolov; then
1 9 l:.a2 .i.g7 ( 1 9 . . .ttJc3 20 �b3) 20
ttJbl bS ! + still favours Black, since
White will have to give back the ex­
change to stop the queenside pawns.
c) I S as is an important move:
older move, also satisfactory) 1 7 �b3
i..xc4 (most often played, although
17 . . .ttJa7 and 1 7 . . .�d7 are also con­
sidered equal, an example of the latter
move being 1 8 l:.d 1 .i.xc4 1 9 �xc4
ttJa7 20 e3 ttJbS 2 1 .td2, Quinn-Sher,
Hastings 1 995/6, and now 2 l .. .b6 22
axb6 l:.xb6 is fine) 1 8 �xc4 ttJd7 1 9
�d3 ( 1 9 e4 ttJeS { easier than 1 9 . . .bS }
20 �e2 c4; 1 9 ttJe4 bS 20 axb6 ttJdxb6
2 1 �d3 c4 =; 1 9 f4 ! ? bS 20 axb6
ttJcxb6 2 1 �d3 c4 22 'ii'f3 ttJcS "with
 FIANCHETTO SYSTEMS WITH g3
double-edged play" - Schneider; 1 9
l:.a3 CiJe7 ! ? 20 �a2 CiJf5 2 1 e 3 CiJe5 22
::i.d l �c7 = Rubinetti-Gheorghiu, Bue­
nos Aires 1 979) 19 ... CiJe5 (or 19 .. :�c7
20 �c2 CiJa7 2 l l:ta2 b5 22 axb6 'iVxb6
= V.Mikhalevski-Ribshtein, Israeli
League 2000) 20 'i¥c2 CiJa7 21 l:Ld l
CiJb5 22 CiJxb5? ! (22 e3 =) 22 ... axb5 23
�3 �d7 24 .i.n CiJc4 ! with the ini­
tiative, Liberzon-Yusupov, Lone Pine
1 98 1 .
15 :�c7 (D)
••
This move-order gives Black an ex­
tra possibility (see the note to Black's
16th move). Otherwise:
a) 15 ... CiJc8 16 'iid3 'iic7 normally
transposes to the main line after 1 7
�e3, since other moves favour Black,
such as 1 7 l:tb l c4 ! , Ljubojevic-Hulak,
Yugoslavia 1 98 1 , intending 1 8 'iixc4
�xc4 1 9 CiJxc4 b5 20 axb5 axb5 2 1
CiJd2 b4 22 CiJa2 .i.b5 +.
b) For once, 15 ... CiJxa4?! seems to
come up short: 16 CiJxa4 CiJxe4 ( 1 6...b5?
17 CiJxc5 ! { or 17 e5 ! ? } 17 ... dxc5 1 8
::i.e l b4 1 9 CiJbl .i.b5 20 e5 ±) 1 7 'iic2
b5 18 .i.xe4 .i.xh3 19 .i.g2 .i.xg2 20
�xg2 bxa4 2 1 CiJc4 :to
121
16 i..e3
1 6 l:tb 1 ! ? CiJxa4 1 7 CiJxa4 b5 1 8
CiJxc5 �xc5 1 9 .i.e3 �c8 20 l:tc 1 'iWd8
+ VMikhalevski-de Firmian, Copen­
hagen 2000. Black has active pieces
and pressure against e4.
16 CiJc8
.••
I think that Black is well enough off
in what follows, but an alternative is
16 . . . CiJh5 ! ? 17 a5 ( 1 7 g4 .i.xc3! 1 8
bxc3 .i.xa4) 1 7 ... CiJc8 1 8 CiJc4 .i.b5
( 1 8 ... b5 1 9 axb6 CiJxb6) 1 9 'iib3 CiJa7
20 �fel .i.xc4 2 1 'i¥xc4 �d7 22 b4
CiJb5 ! with unclear play, Lacrosse­
Ljubojevic, Antwerp 1 994.
17 'iVd3 CiJa7
I don't trust 1 7 ...b5 ! ? at all, a key
game being 1 8 axb5 c4 1 9 'i¥xc4 (or
1 9 b6 :t Osnos; then 1 9 . . . CiJxb6 20
'iie2 intending x:tfc 1 looks difficult for
Black) 1 9 ...'iixc4 20 CiJxc4 axb5 21
CiJd2 b4 22 CiJd l .i.b5 23 l:te1 with an
extra pawn, Quinteros-Gheorghiu,
Novi Sad 1 982.
18 l:Xfcl b5 19 b4
There have been at least six games
from this position, and in none of
them did White try 19 axb5, perhaps
due to 1 9 ...CiJxb5 20 CiJaxb5 .i.xb5 ! ?
(20. . .axb5 2 1 b4) 2 1 CiJxb5 axb5 22 b4
c4 with ...�a8 and (upon .i.d4) ... CiJd7
to follow. Such positions tend to be
satisfactory as long as Black doesn't
have a bad bishop on d7.
19 c4 20 'iifl
..•
Or:
a) 20 �e2 bxa4 ! 2 1 CiJxc4 ! ? CiJb5
22 CiJxb5 .i.xb5 23 e5 dxe5 24 �a2
.i.f8 +. Then 25 .i.c5 ! ? .i.xc5 26 bxc5
�xc5 27 CiJxe5 �d4 28 CiJc6 .i.xc6 29
lId 1 i..xd5 ! was much better for Black
122 THE GAMBIT GUIDE
in the game Korchnoi-Franco, Lu­
cerne OL 1 982.
b) 20 'iWd2 bxa4 2 1 liJxc4 (Birn­
boim-Y.Griinfeld, Israeli Ch 1 986)
2 1 . .Jhb4 =.
20 bxa4
.•.
As seen in the last note, this is the
standard way to activate Black's rook
and clear b5.
21 liJxc4 (D)
21. liJb5!
•.
This appears to be the strongest
move. 2 1 ..Jhb4 ! ? is unclear: 22 liJa2 !
(22 liJxd6 'fWxd6 23 i.xa7 liJxe4 ! 24
liJxe4 :bxe4 25 i.c5 �xd5 26 i.xe4
'iWxe4 =+= Quinteros-Ki.Georgiev, Thes­
saloniki OL 1 984) 22. . ..l:!.b7 23 e5 dxe5
24 d6 'iWb8 25 i.xb7 �xb7 26 liJa5
'iWa8 with compensation - Georgiev.
This is a hard position to assess.
22 liJxa4 liJxe4 23 i.xe4 i.xal 24
liJd2 i.xh3! (D)
Or 24...i.c3 25 'iWd3 (25 �h2 i.xd2!
26 .l:!.xc7 liJxc7 27 i.xd2 l:txe4 with
good play, Glek-Belinkov, COIT. 1 986)
25 .. J�bc8 26 liJb6 f5 (26 .. JIcd8 is un­
clear, according to Glek and Piskov;
TO THE MODERN BENONI
26 . . . i.xh3 ! ? also looks interesting,
with the idea 27 liJxc8 1hc8 ! ? 28 liJb l
'iWe7) 27 liJxc8 lhc8 (27 . . .fxe4 28
liJxe4 ':'xc8 = ECO) 28 i.g2 'fWd8 29
.l:!.b l , Ginting-Lukov, Thessaloniki OL
1 988, and now Ginting mentions the
logical 29 ...�f6. Given the suggested
improvements for Black, 24 . . . i.c3
looks at least equal. But the text-move
(24. . . i.xh3) may be even better.
25 .l:!.xc7
25 .tg2 is supposed to be an im­
provement, but 25 . . . i.xg2 26 'it>xg2
'fWb7 27 .l:!.xal (Glek and Pigusov) al­
lows 27 ...liJc7 ! (27 . . .'fWxd5+ 28 liJf3
'iWb3 ! ?) 28 liJc3 liJxdS 29 liJxdS 'fWxd5+
30 liJf3 .l:!.xb4 with a clear advantage;
e.g., 3 l 1ha6 .l:!.eb8 32 .l:!.al 'iWe4 + with
the idea . . ..l:!.b2, among others.
After the text-move (25 .l:!.xc7), Cvi­
tan-Cebalo, Yugoslav Ch (Novi Sad)
1 985 continued 25 ... i.xfl 26 l:Ic l , and
instead of the game's 26 . . . .l:.xe4 ! ?,
26 ....td4 looks like a safe continua­
tion: 27 �xfl (27 .txd4 liJxd4 28
�xfl .l:!.xb4) 27 ...i.xe3 28 fxe3 f5 29
.tg2 ':xe3 intending ....l:!.xg3 or . . J�a3.
 FIANCHETTO SYSTEMS WITH g3
In general, Black seems to do very
well indeed in this main line with
1 3 . . .tLIb6.
A32)
1l tLIhS (D)
•••
A completely different strategy.
Black prevents f4, prepares ...tLIeS af­
ter tLIc4, and meets e4 tactically. This
is either a second or even third option
if you don't like Line A l or the ... 1:teS
ideas covered in Line B .
12 tLIde4
Considered best by theory. Here are
the alternatives:
a) 1 2 tLIce4 is rather like the text­
move; e.g., 1 2 ...tLIdf6 1 3 tLIxf6+ �xf6!
(this clears the back rank - Black has
no need to fear White's tLIc4-b6; in­
stead, 1 3 ...i.xf6 14 tLIc4 makes it hard
for Black to develop smoothly) 14 tLIc4
and now:
a l ) 14 ...1:tbS I S t2Jb6 ( 1 S as i.d7
16 l::tb l , Derieux-Kovacevic, Massy
1992, and now the most direct strategy
is 1 6 ... i.bS 1 7 b3 'fie7 I S i.b2 i.xb2
19 ':xb2 fS) I s . . :iVdS 1 6 'iWb3 (we've
123
already seen examples of positions
such as 1 6 tLIxcS 1:txcS =, when the
straightforward plan is . . .l:.eS and
...tLIf6, but Black can also contemplate
. . .fS and ... tLIf6, since White can't get
a knight to e6) 1 6...l:!.eS 17 1:te 1 tLIf6
( 1 7 ...fS ! ?) I S i.f4 t2JhS 1 9 i.d2 tLIf6
20 i.f4 1/2-1/2 Reefschlager-Kinder­
mann, Bundesliga 1 9S4/S. If 20 tLIxcS,
then 20 .. :i!Vxcs 21 e4 t2Jd7 22 i.c3 c4
23 'fic2 tLIeS, etc.
a2) Still better looks 14 ... i.d7 ! ? I S
i.d2 .l:.abS (or I S . . .'iWe7 1 6 e4 bS 1 7
t2Jb6 l:.aeS) 1 6 i.c3 'iWe7 1 7 tLIb6 fS !
= with the idea ...tLIf6, Birnboim-de
Firmian, Thessaloniki OL 1 9S4.
b) 12 e4 tLIeS 1 3 'iie2 fS ! = keeps
the balance; e.g., 14 f4 t2Jg4 I S tLIf3 ! ?
fxe4 1 6 'iVxe4? .l:teS 1 7 'iic4 b S +=
Bender-Bogut, Pula 1 996.
c) 12 h3 1:tbS 1 3 tLIce4 tLIdf6 1 4
tLIxf6+ tLIxf6 I S tLIc4 l:.eS 1 6 as tLIe4! =
De Boer-Grooten, Leeuwarden 1 997.
d) 12 tLIc4 tLIeS 13 tLIe3 ( 1 3 tLIxeS
i.xeS 14 f4 i.g7 I S e4 i.d7 = or
I S ... l:.eS =) 1 3 ...l:.bS 1 4 as bS I S axb6
l:.xb6 1 6 h3 l:.b4 1 7 tLIc2 and now, in­
stead of the brilliant but speCUlative
17 . . . fS ! ?, as in Youngworth-de Fir­
mian, Lone Pine 1 9S 1 , Black can just
play 17 ... .l:.bS with equality.
12 :iWe7
••
Or 1 2 . . .tLIdf6 ! ?, intending to an­
swer 1 3 i.gS (or 1 3 tLIxf6+ t2Jxf6)
with 1 3 ... h6.
13 i.gS f6 14 i.d2
14 i.h4 tLIeS is comfortable for
Black.
14 t2JeS 15 'fib3 tLIf7!
•••
This covers d6 and gS, in prepara­
tion for ... fS .
124 THE GAMBIT GUIDE TO THE MODERN BENONI
16 f4 f5 17 lDf2
Now, instead of 1 7 .. :flc7? I S e4, as
in Osnos-Commons, Plovdiv 1 9S2,
Schneider correctly suggests...
17 lDf6! (D)
•..
w play ...lDe4, when after lDxe4 and,.
18 a5
Schneider's move. Much more crit­
ical would be I S e4 and, leaving out
numerous space-consuming details, I
would suggest l S ... lDxe4 ( 1 S ...fxe4 ! ?
might lead to 1 9 l:tfe 1 'Wic7 2 0 lDfxe4
i.f5 2 1 lDxf6+ i.xf6 22 lDe4 i.d4+
23 i.e3 i.g7 =) 1 9 lDcxe4 fxe4 20
:fe1 i.f5 2 1 lDxe4 i.d4+ 22 �h 1
�d7 23 i.c3 i.xe4 24 lhe4 i.xc3 25
�xc3 l:.aeS = 26 J::tae1 (26 l:.xeS 'uxeS
27 a5 lDh6 =) 26 ...'uxe4 27 l:.xe4 b5
2S axb5 axb5 29 'Wif6 l:.aS ! 30 l:.e1 (30
:e7 �f5) 30 ... c4 3 1 i.f3 (3 1 .l:.e7
�f5) 3 l . ..'iYf5 32 'Wixf5 gxf5 =, due to
the queens ide pawns and ideas like
. . J�a2.
18 .:.e8
..•
Now:
a) 19 ':'fe1 'iic7 20 e4 c4 (20...i.d7
is unclear) 2 1 'iic2 fxe4 22 lDcxe4
i.f5 =.
b) After 19 e3, Schneider suggests
19 ... i.d7 20 l:tfe 1 l:.abS 2 1 e4 'iidS ! ,
when Black should be fine.
B) Lines with . . . .::te8
In this section, we look at ...l:.eS ideas
which avoid the main ... lDbd7 systems
of Line A. In most cases, Black will
....:.xe4, the rook is active but also ex­
posed on its 5th rank. As is common
with a second repertoire choice, these
lines tend to be slightly less analysed
and are rather committal, in that Black
is often in a situation where he has ac­
tive pieces, but may have to sacrifice
the exchange to keep his initiative
alive. A general feature is that Black
unleashes his g7-bishop and achieves
queenside pressure.
The reader is offered the choice of
two approaches involving ... l:.eS. They
are obviously closely related, but in­
volve some nuances that can be impor­
tant; ultimately, the decision between
them will probably be a matter of
one's stylistic tastes:
Bl: 9 . l:.e8 . . 1 24
B2: 9 . a6 10 a4 l:.e8
. . 1 27
B1)
9 . l:.e8 (D)
. .
10 i.f4
One has to examine the alternatives
to understand what's going on:
a) 10 l:.e l and then:
a1 ) The easiest answer is 1 0... a6,
when 1 1 a4 transposes to note 'a' to
White's 1 1th move in Line B2, while
1 1 e4 ! ? b5 12 e5 dxe5 1 3 lDxe5 l:.a7 is
equal.
 FIANCHETTO SYSTEMS WITH g3
a2) Black can also try 1 O. ..lLIe4 ! ?
I I lLIxe4 l:he4 1 2 ..tgS 'iVf8 ! 1 3 lL1d2
llg4 14 ..tf4 gS I S ..te3 ( 1 S h3 gxf4 1 6
hxg4 fxg3 1 7 fxg3 ..txb2 1 8 lIbl ..tc3
= ) IS . . . ..txb2 1 6 l:tbl , Razuvaev-Chi­
burdanidze, Tashkent 1 980, and now
16 . . . ..td4 ! ? 1 7 h3 ..txe3 1 8 hxg4 ..td4
19 e3 ..tg7 is an interesting idea to
gain long-term compensation for the
exchange (two bishops, a pawn and
White's poor kingside structure, limit­
ing his bishop). Play might go 20 lLIe4
'fie7 2 1 f4 (after 2 1 f3 lL1d7 or 2 1 'ii'f3
lLId7, Black's position is solid) 2 l . ..h6
22 'ii'f3 lLId7 23 l:tn b6 and Black can
think about queenside expansion or
play ...lLIf6.
b) 10 h3 lL1e4 ! ? (probably easier to
play is 1 O... lLIbd7 1 1 a4 {best, accord­
ing to ECO } , when l l ...lLIe4 1 2 lL1xe4
�xe4 l 3 lL1d2 can now be answered by
l3 ... :e8 =) 1 1 lLIxe4 lhe4 1 2 ..tgS
( 1 2 lLId2 l:te8 =) 12 ...'iVc7 ( 1 2 .. .'iVe8
13 lLId2 1:.b4 has been played) 13 lLId2
l:1e8 14 ..tf4 ( 14 lL1e4 l:txe4 ! I S ..txe4
.txh3 +; fl and b2 hang) 14 ...lLId7 I S
lLIc4 lLIeS 1 6 lLIxeS ..txeS 1 7 .i.xeS
J:LxeS 1 8 e4 ..td7 1 9 'i!i'd2 was given as
125
slightly better for White by ECO. But
the continuation of G.Kuzmin-Bou­
aziz, Riga IZ 1 979, 19 . . . f6 20 f4 l:te7,
was equal according to Schneider,
which seems fair, because Black has
...l:tae8 if necessary, with queenside
play always in the offing.
c) 10 lLId2 is the normal and argu­
ably most important move, which
transposes to Line A3 1 after 1O ... a6 1 1
a4 lL1bd7. Black can try to avoid this as
follows:
c 1 ) 1O ... b6 is quite often played
and probably OK, but it is harder to
play than line 'c2' after 1 1 a4, when
l l . . .lLIa6?! 1 2 lLIc4 slightly favours
White. Black's best line is 1 l .. ...ta6,
when after 1 2 lL1bS, he should play ac­
tively by 12 ...lLIbd7 ! 1 3 lL1c4 ..txbS 14
axbS lLIeS I S lLIxeS ':xeS 1 6 'iVd3,
Panno-Leskovar, Mar del Plata 2000.
In this position, Black can continue
1 6 ...'iVd7 ! 1 7 e4 l:tc8, soon threatening
...c4 and ...'iVxbS, and intending ...l:tee8
and ... l:tc7, the latter move being the
standard method to defend against a
cramping pawn on bS .
c2) 1 O...lLIg4 ! ? (D). Then:
126 THE GAMBIT GUIDE TO THE MODERN BENONI
c2 1 ) 1 1 h3 ttJeS 1 2 f4 ttJed7 1 3 e4
ttJa6 ! ? 14 ttJc4 ttJb6 I S ttJe3 �d7 =.
c22) 1 1 e4 a6 1 2 a4 ttJeS 1 3 'iic2 fS
1 4 h3 b6 I S f4 ttJf7 =.
c23) 1 1 ttJc4 ttJeS 12 ttJxeS �xeS
1 3 'ii'c2 "with the idea l:.b1 , b4" -
ECO; but aside from 1 3 . .. ttJd7 14 :tb1
ttJb6 or 1 3 ...ttJa6, Black even has the
exotic 1 3 . . . �d7 ! ? intending 1 4 a4
ttJa6 or 1 4 :tb1 ? bS.
c24) 1 1 ttJde4 a6 1 2 �gS ( 1 2 a4 fS
1 3 �gS '(Wc7) 1 2 ...f6 1 3 �f4 ttJeS 14
a4 ttJf7 I S ttJd2 gS ! ? ( l S ...ttJd7 =) 1 6
�e3 ttJd7 1 7 ttJc4 ttJdeS 1 8 ttJxeS
ttJxeS 1 9 'ii'c2 fS = Kharitonov-Agza­
mov, Sevastopol 1 986.
If 1O ... ttJg4 (or 1O ... b6) works, and
if Black is satisfied with White's other
10th move options, then this is a good
way to avoid the transposition to Line
A3 1 (9 ... a6 1 0 a4 ttJbd7 1 1 ttJd2 :te8),
which tends to undercut interest in the
...ttJe4 idea.
10 ttJe4 11 ttJxe4 l:.xe4 12 ttJd2
•..
:tb4 (D)
Here 1 2 ... :txf4 ! ? 1 3 gxf4 'ii'f6 ! ?,
Deuster-Grassmeh, COIT. 1 990, should
be met by 1 4 'ii'c 1 ! 'iixf4 I S ttJc4 'ii'f6
1 6 'iie3 �d7 1 7 "fIg3 �f8 1 8 b3 ;1;.
This diagram (see top offollowing
column) looks very similar to the posi­
tion we will see in Line B2 (with ... a6
and a4 in), but there are some different
themes:
13 a3!?
Or:
a) 13 :tb1 gS 14 �e3 �fS IS a3
:txb2 1 6 :txb2 �xb2 1 7 ttJe4 ! ? ( 1 7
'iib3 �f6 1 8 'iixb7 ttJd7 1 9 ttJc4 ttJb6
20 'iia6 ttJxc4 2 1 'ii'xc4 "fIe7 22 �d2
llb8 = Ligterink-Tindall, Hoogeveen
w
1 998) 1 7 ...h6 1 8 1ib3 �xe4 1 9 �xe4
�d4 ! ? 20 �xd4 cxd4 2 1 'iixb7 ttJd7
22 "fIb4 'iif6 23 .l:td 1 l:.b8 24 "fIxd4
'ii'xd4 2S :txd4 1:.b3, Kaidanov-Nor­
wood, Florida 1 993. Black has full
compensation, according to Norwood
(White's bishop is poor, and ...ttJcS is
coming).
b) 1 3 b3 ! ? and now:
b 1 ) 1 3 ... �xa1 ! ? 14 'iWxal was sug­
gested some years ago. White has
good compensation; e.g., 14 ... gS IS a3
:tbS 1 6 �e3 �fS 17 f4! 'iWe7 1 8 'iic3 g4
1 9 �f2 ttJd7 20 e4 �g6 2 1 :tel with
ideas like i.n and ttJc4. This could be
one reason to prefer having ... a6 and
a4 in, but not necessarily so, since ...
b2) 13 ...l:.xf4 1 4 gxf4 i.xa1 looks
safe and is probably best.
13 :txf4! 14 gxf4 �xb2 15 :a2
•••
The point of inserting 1 2 ...:tb4 be­
fore sacrificing on f4: White's rook is
forced to the less favourable a2-square,
as opposed to b 1 .
15. �g7
..
Now:
a) After 16 e4?, 16 ...ttJa6?! (per­
haps lightly ;1;) was played in the game
 FIANCHEITO SYSTEMS WITH g3
Alburt-Peters, USA Ch (South Bend)
1 9S 1 . Norwood, who says that he' s
had the position after l S ... i.g7 " a cou­
ple oftimes", notes that after e4, Black
should play for . . . lbd7-f6-hS with
pressure on the dark squares (a timely
. . . .Jih6 might be handy as well). Thus
1 6 . . .lbd7 ! , and aside from the trip to
hS, Black can also advance on the
queenside, with more than enough
compensation.
b) 16 'iVb3 is less weakening. Then:
b l ) 1 6 . . .lba6 1 7 lbc4 :bS I S a4
lbb4 19 .l:td2 b6 20 .l:tb1 a6 (something
like 20. . .'iif6 2 1 e3 i.fS and . . ..l:teS is
also possible) 2 1 'iia3 was the very
interesting course of Flear-Collas,
French Cht 1 995. Black has systemati­
cally prepared 2 1 . ..bS ! , but he now got
cold feet, probably due to 22 axbS (22
:!.xb4 cxb4 23 'ili'xb4 i.fS ! gives Black
a clear advantage in view of 24 axbS?
l;lxbS or 24 lbe3 as !) 22 . . . axbS 23
l':txb4 (23 lbaS l:taS; 23 lbe3 i.h6)
23 . . . cxM 24 11i'xb4, but then 24. . J:taS !
is practically winning: 2S lbe3 'iiaS +
or 2S 'ilVxbS? l:a1 + 26 i.f1 i.h3 27
lbe3 'iUcS -+.
b2) Another approach is 1 6 . . .lbd7
followed by . . .lbf6 and ...l:tbS. After 1 7
lbc4, even the ambitionless 1 7 ...lbb6 ! ?
I s lbxb6 'ilVxb6 1 9 'ilVxb6 axb6 20 .l:tb1
.l:!.a6 and . . . i.fS looks playable !
In all these lines, Black (with no
weaknesses) has the bishop-pair and a
pawn for the exchange which, along
with White's immobile pawns, ensure
him lasting compensation.
82)
9 a6 10 a4 Ite8 (D)
..•
127
W
A slightly different version of the
. . ..l:teS/. . .lbe4 idea. This time, a black
rook on e4 can swing over to M with­
out being harassed by a3. But the
trade-offs are hard to assess.
11 i.f4
Instead, 1 1 lbd2 again transposes to
Line A3 1 after l l . . .lbbd7, but Black
can deviate (see line 'd'), while the
text-move ( 1 1 i.f4) is the most impor­
tant alternative (compare the analo­
gous line without . . . a6 and a4, i.e. Line
B 1 ) . Still, as in Line B I , White's other
alternatives are noteworthy:
a) 1 1 .l:te1 lbe4 (after l l . . .lbbd7,
1 2 e4 transposes to Line A I , 1 2 lbd2
reaches note 'd' to White's 1 2th move
in Line A3 1 and 1 2 h3 is note 'b3' to
White's 1 1 th move in Line A) 1 2
lbxe4 .l:txe4 1 3 11i'c2 .l:teS 1 4 e4 lbd7 I S
i.f4 fie7 ! (better, I think, than ECO's
l S .. :�c7 1 6 l:tad 1 intending b4) 16 as
.l:tbS =.
b) 1 1 'ilVc2 fic7 ( 1 l ...lbbd7 1 2 e4
'fiIc7 comes to the same thing; 1 1 . ..b6!?
1 2 e4 .l:ta7 is an alternative plan; for
example, 1 3 i.f4 .l:tae7 1 4 lbd2 lbhS,
with the point IS i.gS f6 16 i.e3 fS !)
128 THE GAMBIT GUIDE TO THE MODERN BENONI
12 e4 lLlbd7 1 3 ..tf4, Kuzmin-Aseev,
Tashkent 1 984, and now a natural idea
is 1 3 ...lLlg4 14 as ( l 4 lLld2 l:i.b8 15 h3
lLlgeS 16 ..te3 bS =) 14 . . . lLlge5 1 5
lLld2 11b8 = with . . .bS next.
c) 1 1 h3 b6 ! ? ( l 1 .. .lLlbd7 trans­
poses to note 'b' to White's 1 1 th move
in Line A; 1 l .. .'iilc7 1 2 lLld2 lLlbd7 1 3
lLlc4 lLleS 1 4 lLla3 appeared in Kali­
nin-Juarez Flores, Dubai OL 1 986,
when 14 ... b6 was perhaps best; e.g.,
I S f4 lLled7 1 6 lLlc4 lLlhS 1 7 �h2 fS)
1 2 .l:.el l:.a7 1 3 lLld2 l:.b7 14 e4 b5 I S
axb5 axbS 1 6 lLln b4 1 7 lLlbi l:.be7 ! =
Freeman-Levi, Melbourne 1 996.
d) I 1 lLld2 is critical, as usual, when
1 1 . . .lLlbd7 again transposes to Line
A3 1 . Black can try to avoid this by
l l . ..lLlg4 !?, as in Line B 1 . Play might
go 12 as ( 1 2 lLlc4 lLleS ; 1 2 h3 lLleS 1 3
f4 lLled7 1 4 lDc4 lLlb6 I S lLla3 ..td7 1 6
aS lLlc8 1 7 lLlc4 .i.bS =) 1 2 ... bS ! ? 1 3
axb6 lLld7 1 4 lLla4 lLlxb6 1 5 lLlxb6
'iilxb6 1 6 lLlc4 'iilc7 1 7 .i.f4 lLleS 1 8
lLlxeS .i.xe5 1 9 .i.xeS l:.xeS = with the
idea ...l:.b8.
1l lLle4
•••
l l ...lLlhS 1 2 .i.gS 'ikc7 is very com-
plex and unresolved.
12 lLlxe4 l:txe4 (D)
13 lLld2
Gauglitz-Sandkamp, Solingen 1 998
saw 13 .:ta2 ! ? .i.g4 ! ? 14 b3 .i.xf3 I S
.i.xf3 l:.e8 = .
13...l:.h4 14 .l:.a2
This simple protection of b2 proba­
bly presents Black with the most diffi­
culties. The other move is 1 4 lLle4 h6
and now:
a) 1 5 l:.a2 i.h3 ! ( l 5 ... ..tfS ! ? 1 6
lLlxd6 l:txf4 1 7 lLlxb7 'iilb6 1 8 gxf4
�xb7 with a mess, A.Hoffman-Andres,
La Carlota 1 995) 1 6 ..txh3 l:.xe4 1 7 as
�c7 1 8 �c2, Adorjan-Armas, Bundes­
liga 1 989190, and now Armas gives
1 8 .. J::!.d4 19 .i.g2 bS with an equal po­
sition.
b) I S ..td2 J:txb2 16 'iilc l l:.xd2 1 7
'iWxd2 fS ! ? (perhaps just 1 7. . ..i.xa l ! ?
1 8 11xal ..tfS ! ?; for example, 1 9 'iilxh6
.i.xe4 20 i.xe4 lLld7 2 1 .i.xg6 fxg6 22
�xg6+ �h8 23 'iilxd6 'iWf6) 18 lbc3
'iilaS 19 l:.ac 1 , Savon-Tal, Dneprope­
trovsk 1 970. Then 1 9 ... lLld7 ! offers
Black a lot of play for his small mate­
rial deficit.
14...g5 15 .i.e3 f5
Taking on b2 is always met by lbc4.
16 lLlf3 h6 17 �c1
17 b3 has been played; Black should
simply develop by 1 7 ...lLld7 ! 1 8 .i.d2
( 1 8 lLle l �b6 1 9 .i.d2 1hb3 20 .i.aS
lib1 !) 1 8 ....l:i.e4 =.
17... .i.d7
I think that a new move is needed
here, and the text seems better to me
than the previously played 17 ...'iilf6 1 8
h4 ! , when 1 8 ... f4? ! 1 9 ..td2! favours
White, and Kapengut's 1 8 ... gxh4 1 9
 FIANCHETTO SYSTEMS WITH g3
ttJxh4 ltxb2 "unclear" is virtually re­
futed by 20 lhb2 'tWxb2 2 1 'tWxb2
..txb2 22 iLxh6 ±. Black has held on
in two games after 1 8 ... g4, but White
has a definite positional advantage
with 1 9 ttJd2 ttJd7 20 ttJc4 b5 2 1 axb5 .
18 'ilVc2
Now 1 8 h4 is less effective: 1 8 . . .g4
(messy is 1 8 ... iLxa4! ? 1 9 hxg5 .tb3
20 ':a3 iLxd5 2 1 'tWd2 iLe4) 19 ttJe l
(maintaining the threat on h6; this
time, 1 9 ttJd2 ':xa4 gives White insuf­
ficient compensation) 1 9 .. :iVf6 20 a5
�xb2 = (or 20... .tb5 =) .
After the text-move ( 1 8 'tWc2), play
might continue 1 8 .. :iVf6 19 .td2 ':xb2
20 �xb2 'tWxb2 2 1 'iWxb2 iLxb2 with at
129
least equality, in view of 22 l:tb 1 iLg7
23 ':xb7 iLxa4, etc.
The Fianchetto System can be an
effective weapon in the hands of a
strong player versus an opponent with
inferior positional skills. Nevertheless,
Black has few problems attaining
equality if he knows the theory. The
variations with . . . a6, . . .ttJbd7 and
. . ..:e8 are sound and well-established;
they require some skill in complex
manoeuvring. The variations with an
early ....:e8 and ...ttJe4 (with or with­
out ... a6) have been underestimated, in
my opinion; they should be strongly
considered by players with tactical in­
clinations who like a direct solution.
7 Systems with �d3 a nd
ttJge2
1 d4 lbf6 2 c4 c5 3 d5 e6 4 lbc3 exd5 5
cxd5 d6 6 e4 g6 7 .i.d3
This introduces a very complex and
dangerous system involving .i.d3,
lbge2, 0-0, followed by lbg3, h3 and
f4, along with other ideas such as a
well-timed i.gS. Black has numerous
methods of proceeding which are
held to be sufficient (see the note to 9
0-0), but theory is very confused here
for both sides, with one author sug­
gesting what another rejects. I should
mention that Rainer Knaak is a great
expert on the white side of this system,
and has won several beautiful games
with it.
Out of all systems apart from the
Taimanov Attack in Chapter S and the
Modem Main Line in Chapter 9, this
is probably the one that is the most in­
timidating for Black. Essentially, this
is due to the fact that White threatens
to blow Black off the board on the
kingside via eS and/or fS, and yet keeps
a very firm grip on the centre (note the
overprotection of the king's pawn by
both knights and the d3-bishop). This
means that a pawn sacrifice like . . . bS
will be futile, since it fails to draw
away enough defenders from the cen­
tre. White's pieces are also perfectly
placed to implement the standard
pawn sacrifice ' l ' eS dxeS '2' fS, an
idea we saw in Chapter S. On the other
hand, none of White's pieces are
aimed at eS, and White's basic set-up
requires quite a bit of time to construct
(see lbge2-g3, f4 and h3). With extra
time to initiate counterplay at his dis­
posal, there are features of White' s
position that give hope to the de­
fender. One is that the move . . .c4, sup­
ported by . . :ilc7, is hard to prevent
(the customary transfer lbd2-c4 is not
an option), and it will gain a crucial
tempo on the d3-bishop. Then, owing
to the lack of support for White's eS
attack, Black' s queen knight can rush
can rush to occupy cS and exert influ­
ence all over the board. Another sign
of hope for Black is that it will take so
much time to develop White's pieces
that after . . . c4, Black' s ... bS will be
achieved quickly enough to distract
White from the kingside. In this battle
of ideas, timing is everything.
Before continuing, let's look at an­
other lbge2 system, i.e., 7 lbge2 .i.g7
8 lbg3 0-0 9 .i.e2 (D) (9 .i.d3 is dis­
cussed in the note to White's 9th move
in the main line).
This position arises from the King's
Indian Defence after I d4 lbf6 2 c4
g6 3 lbc3 .i.g7 4 e4 d6 S lbge2 (the
 SYSTEMS WITH i.d3 AND tiJge2
B i.a6 led to equality. But more consis­
'Kramer System' or 'Hungarian At­
tack' ) S . . .O-O 6 liJg3 cS 7 dS e6 S i.e2
exdS 9 cxdS . I consider this properly a
King' s Indian, and indeed analysed it
in my book The Unconventional King 's
Indian. It is considered innocuous, as
indicated by the fact that White tends
to play 9 exdS in the KID move-order,
although that is only equal. Despite
ECO's classification of this position
as a Benoni, 7 liJge2 appears in neither
Psakhis nor Schneider. NCO relegates
it to a short note, and MCO doesn' t
have i t at all. I will therefore give only
an abbreviated overview of two solid
answers:
a) 9 . . . liJa6 1 0 0-0 ( 1 0 h4 is the
usual point of this system, but hS isn't
much of a threat: 1O ... liJc7 1 1 hS bS !
1 2 a4 ! ? { 1 2 i.xbS liJxbS 1 3 liJxbS
llbS 14 liJc3 i.a6 with two bishops
and a lot of compensation } 1 2... b4 1 3
liJbl l:.eS 1 4 h6 i.hS I S i.gS i.a6! 1 6
i.. xa6 liJxa6 1 7 �f3 �e7 { or 1 7 . . .c4 }
1 8 liJd2 �eS 1 9 i.xf6 �xf6 + EPor­
tisch-Timman, Vilnius 1 969) and now:
a l ) 1 O... liJc7 1 1 a4 l:.bS 1 2 l:.bl .
This prepares to answer ... bS with b4.
131
In the game Szabo-Borik, Dortmund
1 974, 1 2...b6 ! ? 1 3 i.f4 l:.eS 1 4 �d2
tent is 12 ...a6 1 3 i.f4 bS 14 axbS liJxbS
(or 1 4 . . . axbS I S b4 cxb4 1 6 lhb4
liJhS ! =) I S liJxbS axbS 1 6 b4 c4 with
equality. White's knight is three moves
from d4, and Black can capture the a­
file.
a2) 1O . . ..:eS 1 1 i.f4 liJc7 1 2 a4
l:.bS 1 3 �d2 a6 1 4 i.h6 ( ' ! ' Forintos
and Haag) 1 4. . .bS I S i.xg7 �xg7 1 6
axbS axbS, Hanks-Kraidman, Gries­
kirchen 1 9S5, and now Forintos gives
1 7 �f4, but among other moves,
Black has 1 7 . . .l:.eS !? (this protects d6
and threatens ...gS with . . . M, when a
centre pawn falls) I S f3 ( I S l:.a7? b4
1 9 liJbl gS 20 �f3 i.g4 2 1 �d3 .i.xe2
22 �xe2 liJcxdS) I S . . .i.d7 =, when
moves like ...b4 and ... liJbS or ... .i.bS
can follow.
b) 9 ... a6 (best according to Forin­
tos) 1 0 a4 liJbd7 is also solid: 1 1 0-0
':eS ( 1 l .. .hS 1 2 i.gS l:.eS 1 3 �d2
�aS ! 1 4 l:.ad l �b4 I S 'ti'c l c4 1 6 f4
liJcS + Jakab-Yu Mingyuan, Budapest
2000) 1 2 i.f4 ( 1 2 �hl l:.bS 1 3 i.gS h6
14 .i.e3 hS IS .i.gS 'iib6 1 6 'ili'd2 liJh7
= Garmendez-Browne, Linares (Mex­
ico) 1 992; 1 2 i.gS h6 1 3 i.e3 l:.bS =
Shemeakin-Moskalenko, Yalta 1 995)
12 ... hS ! 13 i.gS ( 1 3 i.xd6 h4 1 4 liJh l
�b6 +) 1 3 . . .�aS ! ( 1 3 . . .liJfS 1 4 f3
liJSh7 = has also been played) 1 4 'ili'd2
liJh7 I S i.h4 �b4 1 6 ':ad l liJeS
( 1 6 . . .liJb6 ! ) 1 7 �h l liJc4 I S i.xc4
�xc4 with two bishops and activity,
Jakab-MaJada, Budapest 2000.
We now return to the position after
7 i.d3 (D):
132 THE GAMBIT GUIDE TO THE MODERN BENONI
7 i. g7
•••
7 ... a6 8 a4 'iWc7 ! ? (or 8 . . .'iWaS ! ? 9
i.d2 'ifc7) is an attempt to avoid the
transposition given in the next note.
The idea is to play a system with an
early . . . c4; e.g., 9 h3 (9 liJf3 i.g4)
9 . . . liJbd7 1 0 liJf3 c4! ? Black would
probably have to sacrifice the c-pawn
for activity in some lines. This doesn't
really appeal to me, but it could be in­
vestigated. See the next note.
8 liJge2
Here the only serious alternative is
8 h3 (8 liJge2 0-0 9 h3 is analysed in
the next note), which introduces an­
other move-order issue that has been
ignored by theoreticians. Now Black
has no sound way to avoid the Modem
Main Line (Chapter 9) which results
from 8 . . . 0-0 9 liJf3 . Although I've
never seen White actually play 8 h3
and 9 liJf3 in this position, one should
be aware of the possibility.
Of course, those who wait for White
to play liJf3 before going in for the
Benoni don't have to worry about any
of the variations analysed in this chap­
ter.
The only other independent move
after 8 h3 0-0 is 9 i.gS, to which Black
may respond by 9 . . .h6 1 0 i.e3 ( 1 0
i.h4 i.d7 ! ? 1 1 a4 liJa6 1 2 liJf3 'iWb6
1 3 'iWd2 liJb4 14 0-0 liJxd3 I S 'iWxd3
gS 1 6 i.g3 liJhS is analysis by Goldin;
Black has ideas of ...fS ! and . . .'iWxb2)
1O .. J:te8 l l liJge2 ( l l liJf3 c4!? { 1 l ...a6
1 2 a4 liJbd7 is seen more often } 1 2
i.c2 bS) 1 l ...liJbd7 1 2 liJg3 a6 1 3 a4
liJeS 1 4 i.e2 ( 1 4 i.c2 liJc4) 1 4. . .liJh7
I S 0-0 'iWh4 and now, instead of 1 6
liJh l ? ! (Yermolinsky-Sherzer, USA
Ch (Durango) 1 992) 16 ... fS +, Yermo­
linsky suggests that 1 6 'iWel fS 1 7 f4
liJf7 is unclear. See also 8 liJge2 0-0 9
i.gS (note 'b' to White's 9th move).
8 0-0 (D)
•.•
w
9 0-0
Now 9 liJg3 will usually transpose,
but unnecessarily allows Black op­
tions like 9 . . . a6 10 a4 liJg4! ?, intend- '
ing 1 1 h3 liJeS 1 2 i.e2 'iWh4! or 1 1 f3
liJeS 1 2 i.e2 'iWh4. Others:
a) 9 f3 transposes to a fairly harm­
less line which will be dealt with in the
next chapter (Line A).
 SYSTEMS WITH i.d3 AND lDge2
b) 9 .i.g5 used to be played fre­
quently, but has several good answers:
b I ) Note that after 9 . . . a6 10 a4,
1O ... l2Jbd7 I I 0-0 transposes to note
'c' to White's l ith move in Line A,
and 1 O. . J�e8 is another route, since 1 1
f4? ! 'iVa5 1 2 0-0 l2Jg4! threatens ...c4,
Szaraz-Lukac, Slovakian Cht 1 995.
b2) Another solution is 9 ... h6 1 0
.Jtf4 (after 1 0 .ih4, 1 O. . .l2Jbd7 i s re­
garded as equal, but also interesting is
1O ...a6 1 1 a4 'iWa5 1 2 f3 l2Jbd7 13 0-0
l2Je5 = Sliwa-Bertok, Krakow 1 959)
1O. . .a6 (Kapengut likes 1O ... b6) I I a4
'iVc7 1 2 'iVd2 �h7 1 3 f3 l2Jbd7 1 4 0-0
l2Je5 ! ? (Schneider), when 1 5 .ixe5
dxe5 1 6 .i.c4 l2Je8 with the idea ...l2Jd6
is at least equal.
c) 9 h3 is an interesting possibility.
It does prevent the 9 ... l2Jg4 system
(see Line B), but in several lines it
commits White to h3 before he'd like.
White's problem is that, as Psakhis
points out, f3 is the ideal answer to a
number of set-ups involving .. J:Ie8
and/or . . .l2Ja6. Having forfeited that
option (h3 does not go with f3), White
not only allows the main lines we dis­
cuss below (e.g., 9 ... :e8 10 0-0 a6 1 1
a4 'iVc7), but he also opens the way for
several other promising plans for
Black. At least one of these is impor­
tant for Black to look at if his reper­
toire depends upon 9 0-0 l2Jg4 (Line
B). The key position arises after 9 h3
l::.e8 1 0 0-0 (D):
Now 10 . . .a6 1 1 a4 l2Jbd7 transposes
to the note to Black's l i th move in
Line A. Independent ideas:
c 1 ) 10 . . .l2Ja6 1 1 l2Jg3 :b8 (here
1 l . ..l2Jc7 1 2 a4 a6 1 3 .ig5 h6 14 .ie3
133
:b8 is a main line from the 9 0-0 l2Ja6
variation, considered equal; see, e.g.,
Kapengut) 12 a4 l2Jb4 13 .i.b1 ( 1 3
.ic4 l2Jd7 ! 1 4 f4 'iWh4! Kapengut)
1 3 ... a6 14 f4 b5 1 5 axb5 axb5 1 6 'iWf3
.i.b7 = 1 7 f5? l2Jd7 ! 1 8 fxg6 fxg6 1 9
'iVf7+ 'iii>h8 20 l2Jxb5 l2Je5 2 1 'iWc7 .i.a6
+ Aaron-Stein, Stockholm IZ 1 962.
c2) 10 ... c4! ? has done well in lim­
ited tests: 1 1 .i.c2 .i.d7 ( l 1 .. .b5 ! ? 1 2
a 3 a6 1 3 .i.e3 l2Jbd7 1 4 l2Jd4 .i.b7 1 5
'iWd2 'iWc7 "with chances for both
sides" - Schneider; this looks OK) 1 2
a4 l2Ja6 1 3 .ig5 l2Jc5 1 4 f4 b5 ! ?
( 1 4. . .h 6 1 5 .ih4 'ikc7; 1 4. . .'ikb6 ! ?) 1 5
.i.xf6 .i.xf6 1 6 axb5 'ikb6 1 7 �h 1
.i.xb5 1 8 l2Jg3 .i.d7 1 9 :bl :ab8,
Nikolac-Ljubojevic, Zagreb 1 975, is
assessed as somewhere between = and
+ by Schneider, which seems fair.
c3) 1 O....i.d7 ! ? transposes to a the­
oretical line which normally begins 9
0-0 .id7: I I .ig5 ( 1 1 l2Jg3 l2Ja6; I I
.i.f4 b5 ! ? 1 2 .ixd6 b4; I I a4 c4 1 2
..tc2 transposes to line 'c2' ) 1 1 ...'ikc7
1 2 'ikd2 c4 1 3 ..tc2 b5 14 a3 l2Ja6 1 5
l2Jg3 b4 = Spassky-Ljubojevic, Manila
IZ 1 976.
134 THE GAMBIT GUIDE TO THE MODERN BENONI
We now return to 9 0-0 (D):
B the alternative move-orders 9 ...l:.eS and
This is the critical position for the
.i.d3/lDge2 variation. Years of prac­
tice demonstrate that Black has suffi­
cient play here, but this is difficult to
demonstrate without a great deal of
specific analysis. In my opinion (and
that of theory), both 9 ... lDa6 and 9 ...a6
ultimately equalize for Black. The ex­
tensive theory on these two moves pre­
cludes a detailed exposition of both.
I've chosen 9 ... a6 since it is the more
dynamic line, and also because I have in
general suggested more ...a6/...lDbd7
lines than ...lDa6 lines in this book, so
that gives us some thematic consis­
tency. For the record, Kapengut thinks
that the rare 9 ... .i.d7 also equalizes,
and he provides the most thorough
analysis of any theorist on that move. I
should warn that none of these re­
sponses is easy, and White has serious
attacking chances in every line.
Finally, as a second system, I have
chosen the adventurous 9 ...lDg4, which
is more fun and requires much less
study than 9 ... a6.
Thus:
A: 9 a6 •.• 1 34
B: 9 lDg4
•.• 142
Line A also features a discussion of
9 . . .lDbd7.
A)
9 a6
•..
Transpositions abound in this varia­
tion. In the two optional move-orders ,
that follow, I will give a few ways to
transpose to the main 9 ... a6 variations,
without many details. This is done so
that you may avoid certain side-varia­
tions (but accept others) on the way to
the main line.
a) 9 ...l:.eS and now:
a l ) 10 f4? ! c4 1 1 .i.c2 lDg4 + and
Black threatens ...'iWb6+ (Kapengut).
a2) 10 h3 transposes to note 'c' to
White's 9th move above.
a3) 10 lDg3 a6 (or 1 O ...lDbd7 in­
tending to meet 1 1 .i.f4 with 1 1 ...lDe5
1 2 .i.b5 lDfd7 1 3 a4 a6 14 .i.e2 h5 ! 1 5
.:tel h4 i s equal - Kapengut) 1 1 a4
lDbd7 and again we are in the main
line.
a4) 1 0 f3 is Psakhis's objection to
this move-order, and it's true that
. . .l:.eS tends to be best when h3 is al­
ready in. Nevertheless, Black's game
is satisfactory after 1O ...a6 1 1 a4 lDbd7
(a scheme related to one in Chapter 8)
12 'ith l ( 1 2 l:.bl fic7 13 'ith l c4 14
.i.c2 lDc5 15 b4 cxb3 16 .i.xb3 lDfd7 =
Spraggett-Reinaldo Castineira, Dos
Hermanas 2000; 12 b3 fic7 1 3 .i.e3
l:!bS = ) 1 2 . . .l:tbS ! 1 3 ':'bl fia5 (or
1 3 ...lDe5 14 b4 cxb4 1 5 lhb4 lDfd7
1 6 .i.c2 fia5 = Breedveld-Maus, Bad
 SYSTEMS WITH iLd3 AND CDge2
Worishofen 199 1 ; there follows ...ttJc5)
14 i.d2 ttJe5 = Arbakov-A.Kuzmin,
Moscow Ch 1 9S9.
b) 9 . . . ttJbd7 also does the trick;
e.g.:
bl) 10 i.f4 is met by 10 .. :iWe7.
b2) After 10 a4, 1 O ... a6 transposes
to the main line, while 1O ...ttJg4 ! ? is
effectively a tempo up on Line B
(9 ...ttJg4), with the extra move ...ttJbd7.
The move 10 a4 weakens b3, as shown
by 1O ... ttJg4 1 1 h3 ttJge5 1 2 i.c2 l:leS
1 3 f4 ttJc4 14 'i*'d3 ttJa5.
b3) Berliner, who advocates i.d3
and ttJge2, dismisses the whole idea of
playing ...ttJbd7 in the main line by
giving 10 ttJg3 ttJe5 (a waste of time;
of course, 1O ...a6 1 1 a4 and 1 1 . ..l:leS or
1 1 .. :Vllic7 is normal) 1 1 i.e2 here, with
the idea f4/e5, referring the reader to
"any good book on this opening". As
far as I can discover, the completely il­
logical 10 . . .ttJe5? is given only one
trivial reference in all of the literature.
b4) 1 0 h3 ':'eS 1 1 ttJg3 ( 1 1 i.f4
tbe5 1 2 i.b5 �d7 = ; 1 1 a4 a6) l l . ..a6
12 a4 "fIc7 again transposes to the
main line.
10 a4 ttJbd7 (D)
Or 1O . . .l:leS. But this move-order
(l0 ...ttJbd7) is more convenient, since
Black may want to play . . .l:lbS and
.. :Wic7 before ...l:leS.
11 h3
This is almost always played. 1 1
tbg3 is a line of the 5 ttJge2 variation
of the King's Indian which can trans­
pose to the text after 1 1 ..:iVc7 1 2 h3,
but Black has alternatives; for exam­
ple, l l . ..h5 1 2 i.g5 'iVc7. Other moves
for White:
135
W
a) 1 1 f3 transposes to a variation in
the next chapter (see Line A there),
where Black has no difficulties.
b) 1 1 f4 'iVc7 (strange but interest­
ing was 1 l .. .'ife7 12 ttJg3 h5 1 3 h3
':'bS 14 ':'el h4 1 5 ttJf1 ttJh5 ! with
double-edged play, Rusj an-Soln, Slo­
venian U- 1 6 Ch 1 992) 1 2 ttJg3 :eS
( 1 2 ...c4 1 3 .tc2 l:lbS is also played),
and here are some abbreviated lines in
which White tries to skip h3:
bl) 1 3 'iVf3 c4 1 4 i.c2 l:lbS I 5 e5 ! ?
dxe5 1 6 f5 e4 ( 1 6 . . . l:lfS ! ? 1 7 i.e3
ttJc5 intending ...ttJd3 - Kapengut) 1 7
ttJgxe4 tbxe4 I S i.xe4 ( 1 8 tbxe4
ttJe5) I S ... ttJe5; for example, 1 9 'ii'g3
gxf5 20 i.xf5 i.xf5 21 l:lxf5 tbd3 !
with a large advantage (22 'i*'xc7??
l:lel + is a mate in six).
b2) 1 3 'itth l c4 (or 1 3 .. J:tb8 =) 1 4
i.c2 ttJc5 ! ? 1 5 'it'e2 ( 1 5 i.e3 ttJcxe4 1 6
ttJcxe4 ttJxe4 1 7 ttJxe4 f5 + ) 1 5 . . .i.d7
( 1 5 . . . ttJb3 ! ?) 1 6 'it'xc4 b5 with good
compensation.
b3) 1 3 'iVe2 ttJb6 (the alternatives
1 3 ...l:tb8 and I 3 ...ttJxe4 1 4 ttJcxe4 f5
have also equalized) 1 4 f5 ! ? c4 1 5
i.bl ttJbxd5 1 6 'it'f3 ttJxc3 1 7 bxc3,
136 THE GAMBIT GUIDE TO THE MODERN BENONI
Ligterink-Payrhuber, Groningenjr Ech
1 968/9, and it' s hard to believe that
White has enough after Kapengut's
suggestion 1 7 ...iLd7 1 8 iLg5 l:tf8.
c) 1 1 iLg5 l:te8 0 1 ...'iWc7, with the
-
idea ...c4, is also good) 1 2 lLlg3 0 2
'ith 1 'ilc7 1 3 'iWd2 l:tb8 14 l:tac 1 c4 1 5
iLc2 b5, and Black's queenside play
was well underway in Razuvaev-Psa­
khis, USSR Ch (Vilnius) 1 98 1 ) 12 ...h6
1 3 .i.f4 lLle5 ( 1 3 ...'ilc7 1 4 'iWc l ! ? h5 ! ­
Gelfand and Kapengut) 14 iLe2 iLd7
1 5 h3 g5 ! ? 1 6 iLxe5 ( 1 6 iLe3 ? ! g4 1 7
hxg4 lLlfxg4 with a n attack - Gelfand
and Psakhis; a sample line would be
1 8 iLf4 'iWh4 1 9 lLlf5 iLxf5 20 exf5
'iWh5 2 1 'ilc2 c4! 22 ':fe l ! lLld3 23
iLxd3 l:.xe1 + 24 lIKel cxd3 25 'iWxd3
'ilhl + ! with a small edge) 1 6....:xe5
1 7 'iWd2 b5 1 8 axb5 axb5 1 9 ':xa8
'iWxa8 20 iLxb5 iLxb5 21 lLlxb5 lLlxe4
22 lLlxe4 l:txe4 23 lLlxd6 l:td4 24 'iWc2
'iWxd5 25 lLlf5 l:td2 26 'ilc 1 l:te2 27
lLlxg7 �xg7 28 'ilc3+ 112-112 Yusupov­
Kasparov, USSR 1 980.
Thus, White achieves nothing spe­
cial by omitting h3.
1l ...'iWc7 (D)
This is the preferred move-order of
most books, although 1 1 . . .':e8 will
usually transpose. In that case, ... 'iWc7
might sometimes be omitted, as in 1 2
f4 ( 1 2 lLlg3 ':b8 1 3 iLg5 "ilic7 1 4 .:tc 1
c4 1 5 iLe2 b 5 with chances for both
sides, Christiansen-Bu Xiangzhi, Rey­
kjavik 2000) 1 2 . . .'iWa5 ! ? 1 3 iLd2 c4 14
iLc2 l:tb8 = Phillips-Reinhard, Ham­
burg 1 993. This could be a nice back­
up idea.
12 lLlg3
Independent ideas:
a) 12 .i.e3 ':b8 1 3 f4 ':e8 14 'iWel ! ?
c 4 1 5 iLc2 lLlc5 1 6 lLlg3 b5 ! 1 7 axb5
axb5 1 8 e5 dxe5 1 9 fxe5 ':xe5 20 .i.f4
and now Black eventually won after
the promising exchange sacrifice
20. . .lLlfd7 in Baginskaite-Z.llic, San
Francisco 2000, but 20 . . . l:txel 2 1
iLxc7 l:txfl + 22 lLlxfl ':b7 + i s easier.
b) 1 2 b3 .:te8 1 3 l:.bl lLle5 1 4 .i.c2
c4 1 5 f3 0 5 b4 lLld3 1 6 iLxd3 cxd3 1 7
'ilxd3 lLlxe4! 1 8 lLlxe4 ':xe4 1 9 "ilixe4
.i.f5 20 'ikh4 iLxbl is equal), Zhukho­
vitsky-Kopylov, Erevan 1 98 1 , and now
1 5 . . .cxb3 ! ? 1 6 .i.xb3 iLd7 (alterna­
tively, 1 6 ...'ilc5+ 1 7 'ith 1 b5 = ) 1 7
iLe3 b5 = Kapengut.
c) 1 2 iLg5 (this move is seldom
promising in the . . . a6 line) 1 2 ....:tb8 1 3
a5 b5 1 4 axb6 l:txb6 with an equal po­
sition.
d) After 1 2 f4, 1 2 ... l:te8 1 3 lLlg3
transposes to the main line, while '
1 2 . . .c4 1 3 iLc2 lLlc5 is an effective al­
ternative.
12 11e8
••.
For those who are less theoretically
inclined, 12 ... l:tb8 has a decent reputa­
tion:
 SYSTEMS WITH i..d3 AND ttJge2
a) 13 f4 c4 14 ii.c2 bS IS axbS axbS
16 ii.e3 b4 1 7 .:!.a7 'fid8 is fine for
Black.
b) 1 3 'fie2 lIe8 14 .i.e3 hS I S f4 c4
16 ii.c2 h4 1 7 lDh 1 bS 1 8 axbS axbS
19 .:!.a7 'fid8 1eaves White's centre ex­
posed down the e-file, giving Black
time for 20 'fiB b4 2 1 lDe2 lIb7 with a
good game.
c) The best move is 1 3 .i.e3. Then
13 ...%!e8 14 f4 c4 IS i.c2 transposes
to the main line, or Black can vary
with 1 3 . . .c4 1 4 i.c2 bS ( 1 4 .. Jle8? I S
i.a7 lta8 1 6 .i.d4 ± ) I S axbS axbS 1 6
l:ta7 'fid8. Then 1 7 f4 lIe8 transposes
to the note to Black's I Sth move, but
probably 1 7 lDce2 intending lDd4 fa­
vours White instead.
13 f4!?
This gives White some options on
the I Sth move, but if those don't ap­
peal, then 1 3 .i.e3 is better, because in
that case, 1 3 ...c4 1 4 .i.c2 ltJcs I S .i.d4
J:tb8 1 6 f4 transposes to Line A2, and
1 3 ....:.b8 1 4 f4 c4 I S i.c2 transposes
to the main line. So 1 3 ii.e3 avoids the
limitations expressed in the note to
Black's 1 4th move below.
13 c4 14 .i.c2 ltb8 (D)
...
After 14 . . .lDcS ! ? I S 'fiB l:1b8, 1 6
i.e3 transposes to Line A l ( 1 6 'fiB),
but notice that Black has bypassed the
important Line A2 ( 1 6 i.d4).
15 .i.e3
Or:
a) I S �h l aims at eS : I s ...lDcs 1 6
e5 !? dxeS 1 7 fxeS lDfd7 1 8 d6 'fic6 1 9
lDh5 ! ? ( 1 9 lDdS lDxeS 20 lDe7+ ':'xe7
21 dxe7 lDcd3 with compensation)
19 ... bS 20 axbS axbS 2 1 lDxg7 i.b7 !
22 'fid2 �xg7 and Black has a slight
137
advantage, Zakharov-Psakhis, Volgo­
grad 1 977.
b) IS 'fiB bS ( 1 S . . .lDcS ! ? 1 6 eS
dxeS 17 fxeS l:1xeS 1 8 .i.f4 lDfd7 is
playable and obscure) 16 axbS axbS
1 7 eS dxeS 1 8 fS e4! Kapengut. Then
there could follow 1 9 lDcxe4 lDxe4 20
lDxe4 lDeS 2 1 'fig3 .i.xfS 22 ':'xfS
gxfS 23 lDf6+ �h8 24 lDxe8 ':'xe8 =.
White's d-pawn can be weak, and
Black is active, compensating for the
bishops.
15 lDc5
••.
The best move, in my opinion. If
you need an alternative, I think that
l S ...bS 1 6 axbS axbS 1 7 ':'a7 'fid8 (D)
may also be adequate.
Here's an overview of the three
most dangerous lines:
a) 18 'fid2 b4 1 9 lDa4 i.b7 ( 1 9 ...hS
20 eS ! ? dxeS 2 1 fS e4! is unclear) 20
i.f2 (20 'fixb4 lDxdS) 20 . . . i.a8 2 1
i.bl hS 22 eS dxe5 23 fS e4 24 fxg6
fxg6 2S ii.d4 ii.xdS ! (Knaak described
this as "leading by force to a lost posi­
tion") 26 ':!'xd7 ! 'fixd7 27 i.xf6 e3 28
'it'c2, Knaak-Balashov, Leipzig 1 973,
and now Fritz finds 28 . . .e2 ! 29 lDxe2
138 THE GAMBIT GUIDE TO THE MODERN BENONI
(29 'iixg6 exfl 'it'+ 30 lbxfl 'iia7+ 3 1
Wh1 :e2 32 'iVh7+ Wf8 -+) 29 ...'iia7+
30 lbd4 (30 Wh2 �f7) 30 ....l:i.e4 3 1
JL.xg7 Wxg7 with a clear advantage for
Black.
b) 1 8 eS dxeS 1 9 fS JL.b7 (another
idea is 1 9 .. J,lb7 ! ?) 20 lbge4 and now
Kapengut proposes 20 . . . b4 ! ? His
line continued 2 1 lbbS(?), but then
2 l .. .lbxe4 ! 22 JL.xe4 lbf6 is just win­
ning for Black. 2 1 lbd6 is better, al­
though 2 l ...JL.a8 ! ? 22 lbxe8 'iVxe8 23
lbe4 lbxdS seems to give Black suffi­
cient compensation.
c) 1 8 'iWf3, and while theory rec­
ommends 1 8 ...b4, I think that simply
1 8 ...'iie7 ! ? is attractive as well; for ex­
ample, 1 9 �c 1 ( 1 9 JL.d4? lbxe4 20
JL.xg7 lbd2; 1 9 lbge2 b4 20 lbd4!
JL.b7 ! ; 19 �d 1 b4 20 lbce2 JL.b7 2 1
�d4 hS ! ?) 1 9 . . .b4 20 lbd 1 (20 lbce2
lbxdS ! ) 20 . . J:tb7 2 1 �a6 1!c7 with
equality.
After the text-move ( 1 S . . . lbcS),
there are two wild variations, both ex­
tremely dangerous (for both sides !):
AI: 16 'iVf3 1 38
A2: 16 JL.d4 140
I don't know if anyone has sug­
gested 16 eS dxeS 17 fxeS. Probably
the centre is too exposed after 17 ...lbfd7
(or 1 7 . . JheS ! ? 1 8 JL.f4 lbfd7 1 9 JL.xeS
lbxeS 20 d6 ! ? 'iWd8, a standard type of
exchange sacrifice for a pawn, with
one knight going to d3 and moves like
. . .JL.e6 and .. :ikh4 in the air; this is
probably just sufficient compensation,
but it would be hard to play White) 1 8 ,
d6 ( 1 8 e6 fxe6 1 9 d6 �d8) 1 8 ...'iic6 "
(or 1 8 ...�d8 1 9 'ikdS lbe6 20 'it'xc4 "
JL.xeS, which can only be considered a
mess) 1 9 lbdS lbxeS 20 lbe7+ ':xe7
2 1 dxe7 JL.d7, again with typical com­
pensation in the form of activity and
prospects of gathering up the e-pawn.
AI)
16 'iWf3 (D)
B
The move approved by most theo- .,
reticians. It is reasonably well worked �
out, but both sides need some neW
ideas, which I hope I've supplied.
16 b5
...
The famous encounter Penrose-Tal, "
Leipzig OL 1 960 illustrates the sort of
 SYSTEMS WITH
position I'd like to avoid: 1 6 . . .lbfd7
1 7 �f2 ! ? b5 1 8 axb5 axb5 1 9 e5 ! dxe5
(not much better was the recent 19 ...b4
20 lbce4 lbxe4 2 1 lbxe4 dxe5 22 f5 of
Salas-Lie, Istanbul OL 2000) 20 f5
.i..b7 2 1 l:tad l ..ia8 22 lbce4 ±. I can­
not emphasize enough how easy it is
to allow this manoeuvre, which under­
lies so much of White's strategy in this
system.
17 axbS axbS 18 eS dxeS 19 fxeS
Ineffective is 1 9 f5 e4! 20 'iVf2 (or
20 lbgxe4 lbcxe4 2 1 lbxe4 lbxd5 !)
20 ...lbd3 2 1 ..ixd3 cxd3. This looks
good for Black, in view of22 ..if4 'ii'b6
23 .i..e3 (23 ..ixb8? e3 ! =+=) 23 ...'iWd8 24
fxg6 fxg6 intending to meet 25 ..ig5?!
with 25 . . .e3 ! .
19 JbeS (D)
..
w Kapengut's main line is 2 1 . ....if5 ! ?
20 .td4
20 .Jtf4 ! ? is seldom seen: 20 ...b4
(Kapengut proposes instead 20... lbfd7
21 .Jtxe5 lbxe5 ! with compensation)
21 ..ixe5 'iVxe5 22 l:hel 'ii'd4+ 23 cJ.Ih l
i.xh3 ! ? 24 lbce4 ! (a clear improve­
ment upon 24 lbce2? 'ii'xb2 25 'ii'f4
�c8 =+= Kasparov-Rachels, New York
iLd3 AND 0,ge2 139
simul 1 988) 24 . . .lbfxe4 ! (I think that
this is better than 24 . . .lbcxe4?! 25
lbxe4 ..if5 26 lbxf6+ 'iVxf6 27 'iWf4
nc8 28 d6 .:td8, Cording-Lobron, Bun­
desliga 1 98 1 12, when 29 ..ixf5 ! gxf5
30 l:tdl is in White's favour) 25 'iixf7+
(25 lbxe4 ..if5 !) 25 . . . cJ.Ih8 26 lbxe4
(26 ..ixe4 ..id7 +) 26.....if5 27 lbf2! (27
lbg5 'iYh4+ 28 lbh3 lbe4) 27 .....id7 28
lbe4 ..if5 = (or 28 ... ..ib5 ! ?).
20 ngS 21 lbge2!
.•.
Other tries such as 2 1 lbge4? lbcxe4
22 ..ixe4 lbg4 -+ Timman-Ljuboje­
vic, Amsterdam 1 975 have failed here.
The only real alternative to the text­
move is 2 1 'iWf2 ':'xg3? ! (Kapengut's
2 1 .. ...ixh3 ! looks very strong) 22 d6
'iVxd6 23 ..ixc5, Paulsen-Kettner, Bun­
desliga 1 98617, and now 23 ...l:txg2+!
24 cJ.Ixg2 ..ib7+ 25 cJ.Igl 'iVc6 26 'iti>h2
lbd7 ! would still favour Black.
21 lbhS! (D)
•••
2 2 'iie3 ..ih6, but I don't fully trust it.
22 ..ixg7
Alternatives are messy but fine for
Black:
140 THE GAMBIT GUIDE TO THE MODERN BENONI
a) 22 .te3 is untried; some ideas:
22 ...l:teS 23 g4 (23 d6 'fIid7 24 .td4
.tb7 +; 23 ltJd4 ltJd3 ! ? 24 ltJdxbS { 24
ltJcxbS 'fIid7 ! 2S g4 �xdS ! ; 24 .txd3
cxd3 2S ltJdxbS 'fie7 26 .ta7 :b7 27
g4 d2 ! } 24 . . .'fie7 ! 2S .ta7 ltfS +)
23 . . .b4 24 gxhS? ! (24 ltJd l ltJf6 2S
.tf4 .tb7 ! ) 24 ...bxc3 2S hxg6 hxg6 26
bxc3 l:tb2 +.
b) 22 d6 'fIid7 23 .txg7 �xg7 24
ltJf4 (24 'fie3? f6 2S l:tadl i.b7 ! -+;
24 ltJdS i.b7 2S 'fic3+ f6 26 ltJdf4
ltJxf4 27 ltJxf4 'iixd6 and Black wins,
Heiling-Kreuzer, Germany tt 1 993/4)
24 . . .ltJg3 ! + (24... 'fixd6 2S ltJxhS+
l:txhS 26 'f1ixf7+ 'it>h6 =) 2S ltJfdS (2S
ltJcdS? 'fIixd6 26 'f1ic3+ 'flieS; 2S l:tfdl
.tb7) 2S ... ltJxfl 26 �f6+ 'it>h6! and
Black has a clear advantage, Toma­
szewski-Panczyk, Polish Ch 1 986.
22 'it>xg7 23 d6 'fixd6
•••
This is how the original game with
this line went. Kapengut suggests
23 ...'fib7 ! ? instead.
24 'fixf7+ �h6 25 ltadl
Inferior is 2S 1Hd i 'fif6 +. 2S 'fif8+
'fixf8 26 l:txf8 b4 27 ltJe4 l:teS is a bit
chaotic, but good for Black.
25...ltJd3!
I think that this is a big improve­
ment over Knaak-Enders, East Ger­
many 1 982, which went 2S ...'fIieS ? 26
h4! . Actually, I think that Knaak's own
idea of 2S ...'fic6 ("unclear") also looks
good; e.g., 26 'fif8+ (26 g4 .tb7)
26. . . ltJg7 27 :f2 (27 l:tdS ltJce6 ! )
2 7...ltJce6 2 8 'fin 'fic7 ! ? 29 ltJe4 ltfS
30 'fixc7 ltJxc7 +.
After the text-move (2S ... ltJd3 0,
play might continue 26 ltJe4 (what
else? Both ....txh3 and ... .tb7 were
threatened, and this time 26 'fIif8+?
fails to 26 . . .'fixf8 27 l:txf8 .txh 3 ! )
2 6...flib6+ 27 <;i;Jh l l:tfS 28 1::txfS .txfS
and Black has an obvious advantage.
It is so difficult to break down the
Benoni by direct attack as long as
Black stays aggressive !
A2)
16 .td4
This is yet another menacing move
which threatens to cast Black's game
into doubt. And here too, Black's play
needs patching up. Still, in the end I
think he is doing well:
16...b5 17 axb5 axb5 IS f5 (D)
Given ' ! ' by most theoreticians. 1 8
eS dxeS 1 9 fxeS l:txeS 20 'fif3 trans­
poses to Line A I .
IS...l:tfS!
I completely disagree with theory
here, as I think that there are two play­
able moves, neither of which is the one
claimed to be best. The maligned text­
move is quite sound, in my opinion,
and I will offer another intriguing idea
for back-up:
 SYSTEMS WITH
a) 1 8 ...g5?! is the theoretical move.
It has some nice games to support it,
but I find it very difficult to fight
against Knaak's discovery 19 'ii'c I ! h6
20 h4! b4 (20...gxh4 2 1 i.xf6 i.xf6 22
'Yi'xh6! i.d4+ 23 �h l hxg3 24 f6
i..xf6 25 1::txf6 +- Knaak; 20 ...tbh7 ! ?
21 ..txg7 ! Wxg7 2 2 hxg5 tbxg5 23
tDh5+!) 2 1 hxg5 bxc3 22 bxc3 ! ? (22
gxf6 appears to be even better in view
of 22 ...cxb2 23 ..txb2 'ii'b6 24 'ii'd 1 ! -
Knaak) 22...tbfxe4 23 tbxe4 hd4+ 24
cxd4 tbxe4 25 gxh6 f6, Knaak-Ban­
giev, German Ch (Bremen) 1 998, and
now Knaak's 26 'ii'e 1 ! 'ii'h7 27 i.xe4
still keeps some advantage.
b) 1 8 ...tbfd7 ! ? is a move no one
mentions, but it is certainly worth ex­
amining. I tried a little experiment
here. Even with a good deal of time,
two analytical engines refused to sug­
gest this move even as a fourth option!
Another engine declared it as the sec­
ond best, but wouldn't promote it, and
a fourth one took its time, but then de­
clared 1 8 . . .tbfd7 the best move ! In one
of those mysterious twists, every en­
gine liked 1 8 ...tbfd7 best (not neces­
sarily correctly) once I had played it
for them. At any rate, this is the sort of
position one could spend a few days
engrossed in. Since 1 8 ...1:.f8 seems to
hold up, I'll just supply some fascinat­
ing but incomplete analysis:
b1 ) 19 fxg6 i.xd4+ (or 1 9...hxg6)
20 'iVxd4 hxg6 =. Black can attack on
the queenside while defending with
moves like ...tbe5 and . . .'ii'e7.
b2) 19 ..txg7 Wxg7 20 'ii'd4+ (20
fxg6 hxg6 21 'ii'f3 tbe5 22 'ilkf6+ Wg8
= 23 tbce2 b4 24 tbd4 b3 25 ..tbl
Ji.d3 AND lbge2 141
tDcd7 26 'ii'f4 c3 !) 20...tbe5 2 1 f6+ (2 1
'ilkd2 b4 22 tbce2 tbcd3 23 f6+ �h8
24 i.xd3 tbxd3) 2 1 ...<ifi>h8 22 'ii'e 3 b4
23 'iib6 :g8 =. Black's queenside at­
tack and d3 outpost make up for his
cramped kingside; for example, 24
tbce2 tbcd7 ! , contemplating ...g5 and
...l:.g6.
b3) 19 f6 i.h8 ( 1 9. . ...tf8 is proba­
bly safer; e.g., 20 'ilkc 1 b4 2 1 tbce2 c3
22 b3 i.a6 ! ?) 20 'ii'd2 (20 tbf5 gxf5 2 1
'ilkh5 l:.e5 ! ? 22 'ilkg5+ <ifi>f8 23 'ii'h6+
�e8 24 'ii'xh7 i.xf6 25 'ilkg8+ We7 +)
20...b4 2 1 tbce2 tbe5 22 tbf5 ! ? i.xf5
23 exf5 c3 24 bxc3 b3 25 fxg6 hxg6
26 i.b1 b2 27 l:.a2 tbb3 followed by
... tbc4 and/or ... tbxd4.
b4) After 1 9 tbce2, truly compel­
ling play can follow 1 9 ...tbe5 ( l 9 ...b4
is an equally obscure possibility) 20 f6
i.h8 (20... i.h6 ! ?) 2 1 'ii'd2 tbcd7 22
'ilkg5 'ii'd8 23 tDh5 ! h6 ! ? (most excit­
ing, but 23 . . .b4 and 23 .....tb7 are not
so forcing) 24 ii'xh6 ! gxh5 25 :a3 !
(25 'ilkg5+ tbg6) 25 . . .tbxf6 26 ':g3+
with beautiful play: 26... i.g4 27 hxg4
hxg4 28 i.xe5 ! 1:.xe5 29 ':xf6 ! i.xf6
30 .:txg4+ i.g5 3 1 tDd4 (3 1 tbf4 'fie7
32 tbh5 f6! with the idea 33 tbxf6+
�f7 !) 3 1 . . .:b7 ! 32 tbf5 (32 tbc6
'ii'b6+ 33 �f1 'ii'e3 ! =) 32 .. Jhf5 33
exf5 f6 34 ':xg5+! fxg5 35 f6 'fib6+
36 'it>h 1 'it'e3 =. Thanks to Hiares and
Fritz for their kind assistance !
We now return to 1 8 ...:f8 (D):
19 'it'c1
This is the move that was supposed
to put 1 8 ...':f8 out of business. Per­
haps 1 9 fxg6 fxg6 20 'ii'd2 is objec­
tively better, but Black has no serious
difficulties after 20... b4; e.g., 2 1 tba4
142 THE GAMBIT GUIDE TO THE MODERN BENONI
b3 (2 l .. .tDcd7 ! ?) 22 .ib l l:ta8 23
.ixc5 dxc5 24 d6 'tWaS ! , etc.
19 b4 20 .ixf6 .ixf6 21 e5
...
This move was played in the well­
known game Knaak-Teske, East Ger­
man Ch (Nordhausen) 1 986, won in
brilliant fashion by White after Black
played 2 1 .. . .ixe5 22 f6 ! .ixf6 23
Ihf6 bxc3 24 'tWh6 ! . But I think that
Black has a real improvement with the
unaesthetic but solid 2 1 .. .dxe5 ! ; e.g.,
22 fxg6 (22 'tWh6 .ig7 23 llVh4 bxc3
24 f6 .ih8 25 tDf5 .ixf5 26 l:txf5
':xb2 -+; 22 tDce4 tDxe4 23 tDxe4
'tWb6+ 24 �h l gxf5 25 tDxf6+ 'iWxf6)
22 ... fxg6 (22 ...bxc3? 23 ':xf6 cxb2?
24 llVh6! mates) 23 tDce4 tDxe4, and
Black is at any rate not worse.
Given Black's many alternatives
along the way, he needn't enter into
the last part of this wild main line
(from move 16 on); on the other hand,
he seems to have plenty of chances for
advantage by doing so. I don't often
suggest a theoretical variation that is
so horribly convoluted as this one, but
the .id3/tDge2 lines are dangerous, and
this line promises Black the chance for
an advantage. At any rate, one could
learn an enormous amount about dy­
namic chess play by carefully study­
ing these lines.
B)
9 tDg4 (D)
...
Perenyi's relatively obscure move,
which Attila Schneider has both played
and investigated in some depth. Inter­
estingly, in his thorough monograph
on A65 (the code which includes this
.id3/tDge2 system), Kapengut assesses
9 ...tDg4 as 'unclear' in all main and
side variations.
What is the point of this move?
First of all, against lines with tDg3,
. ..llVh4 is a handy way to harass the
kingside. If White plays the normal 1 0
h3, Black plays 1O. . .tDe5, and when
White moves his bishop and plays f4,
Black will often reply ...tDc4, and fi­
nally, when his knight is attacked
again, he plays ...tDa5 ! At first, it may
seem absolutely insane to make five
moves with the same knight and end
up apparently misplaced on the edge
 SYSTEMS WITH
of the board on as ! Yet White himself
has played four extra moves, some of
them weakening, to force the knight to
its new post, and in reality, the knight
is often quite useful there. Another
way of looking at it is that the knight
on f6 is a bit of a 'problem' piece in
this line, allowing those eS and fS at­
tacks, masking the bishop, and pre­
venting .. .fS. Apart from ... ttJe8, which
can have its own drawbacks, this is
one of the only moves which addresses
those problems without getting in the
way of the rest of Black' s develop­
ment.
All of this doesn't mean that Black
stands better by any means, but 9 ...ttJg4
certainly shakes things up and creates
interesting counterplay. What's more,
the theory of 9 ...ttJg4 is totally unre­
solved, offering room for fresh ideas
and original play.
lO h3
Certainly the main move, but most
of the themes of this line are illustrated
in the following alternatives:
a) 10 ttJg3 leads to rich play:
a l ) 1O ... a6 ! ? and now:
a 1 1 ) 1 1 .te2 'iWh4 12 .txg4 .txg4
1 3 'iUc2 bS Schneider. Then 14 .tf4
'fie7 I S f3 .tc8 ! ? 1 6 a4 M with ... as
and ... .ta6 is fully equal, as are plans
with ... ttJd7-eS .
a12) 1 1 a4 �h4 1 2 h3 ttJxf2 !
( 1 2 ...ttJeS 1 3 .te2 ttJbd7 is also possi­
ble) 1 3 .l:Ixf2 ( 1 3 �xf2 .td4+ 14 �f3
hS ! intending ... .tg4+) 1 3 .. :ii'xg3 1 4
�f4 Wih4 I S .txd6 ( 1 S ttJe2? "JiIIe7 1 6
'fid2 ttJd7 1 7 ':'afl ttJeS with an essen­
tially winning advantage for Black,
Grom-Tolnai, Velden 1 993) I S ... .td4
i..d3 AND 0,ge2 143
1 6 �f3 ttJd7 ! ( 1 6...'iWxf2+ 1 7 �xf2
.txf2+ 1 8 �xf2 :e8 1 9 .txcS ttJd7 20
.td4 ttJeS =1=) 1 7 1:tfl ttJeS 1 8 .txeS
.txeS 1 9 �h l gS ! and Black is better.
a2) 1O .. :i¥h4 1 1 h3 ttJxf2 ! 1 2 1:txf2
�xg3 1 3 .tf4 �h4 avoids the 1 O...a6
1 1 .te2 idea in line ' a l l ' . Then the
play parallels line 'aI2' unless White
tries 14 ttJbS a6 I S ttJc7 1:ta7 16 .txd6
.td4 1 7 'iWf3 ttJd7 ! 1 8 1:tal ttJeS 1 9
.txeS .txeS 20 ttJe6 .txe6 2 1 fxe6 b S !
+.
b) l O :Ib 1 ttJa6 (stopping b4) 1 1
h3 ( 1 1 a3 ttJeS 1 2 .tc2 c4) 1 1 ...ttJeS
1 2 .tc2 ttJb4 1 3 .ta4 a6! ? ( 1 3 ... fS 1 = )
14 a3 ttJbd3 I S .te3 c4 1 6 f4 ttJd7 1 7
.tc2 �aS 1 8 b4? ! 'iWxa3 1 9 .txd3
.txc3 20 .tc 1 ? 1i.d4+ ! 21 ttJxd4 ir'xd3
+ Effert-Haist, Zell 1 99 1 .
c) 1 0 1i.c2 ttJeS 1 1 b3 bS 1 1 2 :Ibl
( 1 2 f4 ttJg4 1 3 h3 M), Ghitescu-Mol­
dovan, Bucharest 1 995, and now Kap­
engut suggests 1 2 . . . a6 as 'unclear' .
Here 1 2 ...b4 1 3 ttJa4 and 1 3 ...ttJbd7 ! ?
or 1 3 ...1i.a6 looks fine.
d) 10 .tf4 ttJa6 ( 1 O ...fS ! ? 1 1 exfS
.txfS, Morante-Estebanell, Cajas 1 989,
may also be playable, but I don't fully
trust 12 .txfS l::txfS 13 .tg3 with moves
like ttJf4 and �a4 to come) and now:
d 1 ) 1 1 ttJg3? ! ttJxh2 ! 1 2 .txd6
.tg4! 1 3 �b3 'tWxd6 14 �xh2 hS ! I S
�g l h4 1 6 ttJge2 ttJM =1= Purgimon­
Povah, COIT. 1 987-94.
d2) 1 1 a3 �e7 12 :bl ttJeS 1 3 .tc2
.td7 14 b4 cxM I S axM l1fc8 1 6 .ta4
':'c4 ! 1 7 .txd7 ttJxd7 (or 1 7 ...ir'xd7 1 8
bS ttJcS) 1 8 ttJbS ttJeS =1= P.1anse-Bez­
emer, Haarlem 1 999.
d3) 1 1 'iVd2 ttJeS 12 .tc2 ttJc4 1 3
ir'c 1 ttJb4 1 4 ..tbl ttJeS I S 'i¥d l ( ' ? l ' -
144 THE GAMBIT GUIDE TO THE MODERN BENONI
Kapengut) 1 5 ...'iIi'a5 ! 1 6 a3 c4 1 7 'ili'd2
liJbd3 1 8 b4 ! ? 'ili'd8, Miles-Perenyi,
Porz 1 986. In this position both Kap­
engut and Schneider like Black.
d4) l l liJc1 (perhaps best) l 1 ...liJe5
1 2 i.e2 f5 1 3 'ili'd2 liJc7 = 14 liJd3? !
( 1 4 i.g5 ! ?) 1 4...liJc4 1 5 'ili'c 1 b5 ! +,
having in mind 16 liJxb5? liJxb5 1 7
'ii'xc4 liJd4 =+= G.Martin-Sichev, COIT.
1 99 1 .
e) 1 0 f4 ! ? (D) sacrifices an ex­
change for some play, if Black chooses
to accept it. Of the sources I have that
take 9 ...liJg4 seriously, none of them
mention 1 0 f4, probably (perhaps cor­
rectly) assuming that White lacks
compensation.
This position is interesting enough
to warrant fairly detailed coverage of
several possibilities for Black:
e l ) 1O ... c4 turns out to be rather
risky but quite possibly playable; e.g.,
1 1 i.xc4 'ili'b6+ 12 Wh l liJf2+ 1 3
%:'xf2 'ili'xf2 1 4 liJb5 ! (the only try for
compensation) 1 4...'iIi'b6 1 5 'ili'd3 ( 1 5
�b3 a6 1 6 i.e3 'ili'd8 1 7 'ili'a3 comes to
the same thing) 1 5 ... a6 1 6 i.e3 'ili'd8
1 7 'ili'a3 axb5 ! 1 8 'ili'xa8 bxc4 1 9 'ili'xb8
'ili'e7 20 liJg3 i.f5 ! 2 1 'ili'a7 i.xe4 with
at least equality; e.g., 22 i.d4 i.xd4
23 'ili'xd4 :e8 with the idea that 24
'ili'xc4 is met by 24...i.xd5 1 . This may
not be to everyone' s liking, and yet
White cannot be happy groping around
either.
e2) 1 O . . . f5 is an attractive thrust
that can resemble 1O ... c4: 1 1 exf5 ( 1 1
h3 c4 ! ; 1 1 e5 dxe5 1 2 h3 e4 1 3 i.c4
liJf6 =+=) l 1 ...gxf5 ( 1 1 .. .c4! ? 1 2 i.xc4
'ili'b6+ 1 3 Wh l i.xf5 14 'ili'el :e8 is
unclear) 1 2 'ili'b3 (12 h3 c4 13 i.xc4
'ili'b6+ 14 �h l liJf2+ 1 5 :xf2 'ili'xf2 1 6
liJb5 'ili'c5 1 7 'ili'd3 %:.e8; 1 2 'ili'c2 liJa6
1 3 a3 c4 14 i.xc4 'ili'b6+ +) 1 2 ...:e8
1 3 h3 liJh6 = 14 i.d2 a6 1 5 a4 b6 ! ? 1 6
:f3 %:.a7 = intending ...:e7.
e3) 1O . . .'iIi'h4!? 1 1 h3 liJf6 is un­
tidy, but better than it looks, because
the queen is annoying on h4 and the
combination of ...liJd7 and ...liJe8 can
restrain the centre; e.g., 1 2 'ili'el ( 1 2
liJb5 liJe8 1 3 f5 liJd7 = ; 1 2 e5 dxe5 1 3
fxe5 liJfd7 1 4 liJe4 'ili'e7; 1 2 f5 liJe8)
12 .. :iWxel 1 3 :Lxel a6 ! ? (or 1 3 ...liJbd7)
14 a4 ( 1 4 e5 dxe5 15 fxe5 liJfd7 =)
14 ...b6 ! ? 1 5 b3 :La7 16 i.b2 :Le7 =.
The lines in this note are remark­
able, in that perfectly logical play by
White leads nowhere. Hence the forc­
ing 1 0 h3 (D), to which we now return:
lO liJe5 11 i.e2
•.•
Or:
a) 1 1 i.e3?! is easily met by l 1 ...f5
or l l .. .liJxd3 1 2 'ili'xd3 f5.
b) 1 1 i.bl liJa6 (with the bishop on
b l , 1 1 . ..a6 ! ?, which is questionable
with the bishop on c2, may be used to
exploit the b3-square: 1 2 a4 %:.e8 1 3 f4
 SYSTEMS WITH
{ 1 3 b3 'iih4! ? } 1 3 ...lZJc4 1 4 'iid3 { 14
b3 lZJaS IS ..tc2 bS } 14 ...lZJaS I S ..ta2
lZJd7 = with the ideas . . .c4 and/or
... �6; White lost a key tempo trying
to defend b3, and still has a good deal
of trouble with that square) 1 2 f4 lZJc4
13 �d3, Roemer-Perenyi, Balaton­
bereny 1 988, and now Kapengut sug­
gests 1 3 ...bS ! 14 b3 lZJb4 IS 'iif3 lZJb6
1 6 .i.b2 ( 1 6 a3 lZJ4xdS 1 7 exdS b4)
16 .. .fS 1 7 a3 lZJa6. In fact, 17 .. .fxe4 1 8
.i.xe4 lZJa6 i s positionally attractive as
well.
1l lZJa6 12 f4 (D)
.•.
Again the most common move.
Others:
a) 1 2 lZJg3 lZJb4 ! ? ( 1 2 ...'iih4) 1 3
i.bl 'iih4 1 4 a3 lZJa6 I S 'iie2 lZJc7 1 6
.i.e3 bS ! = Tsiganova-Chernikova,
Melitopol 1 992; then Schneider offers
1 7 lZJxbS? ! lZJxbS 1 8 'iixbS ..txh3 ! .
b) 1 2 a3 l:tb8 1 3 f4 lZJc4 14 ..td3 ( 1 4
b3 lZJaS I S fS bS 1 6 .i.b2 .i.eS = Soln­
Wolter, Schwarzach 1 999) 14 . . .lZJaS
(the Perenyi manoeuvre ! Otherwise
14 ... lZJxb2 IS .i.xb2 c4 16 .i.xc4! 'iib6+
17 �hl :;!; Knaak-A.Schneider, Stara
Zagora Z 1 990) I S .i.e3 bS ! ? ( 1 S ...fS
iLd3 AND 0,ge2 145
is quite sensible, intending to meet 1 6
exfS ..txfS 1 7 i.xfS b y 1 7 .. JhfS with
control of the queenside light squares,
or 1 7 ... gxfS) 16 ..txbS ( 1 6 b4 might be
better; still, something like 1 6... cxb4
1 7 axb4 lZJxb4 1 8 i.xa7 l:ta8 1 9 i.d4
..txd4+ 20 lZJxd4 lZJxd3 2 1 'iixd3 b4
probably holds the balance; e.g., 22
lZJcbS ..ta6 23 l:tfbl lZJb7) 1 6... lZJc7 1 7
i.d3 l:txb2 1 8 'iia4 and now, apart
from 1 8 ...lZJa8 1 9 l:tfbl .i.d7, which
was unclear in Urban-StajCic, Mied­
zybrodzie 1 99 1 , 1 8 ... c4! merits atten­
tion.
12 lZJc4
...
This was Perenyi 's original idea,
but as an alternative, 12 ...lZJd7 looks
fully playable as well:
a) 1 3 lZJg3 l:tb8 1 4 �h l bS (or
14 ...lZJc7 ! ? I S a4 a6) I S a4 bxa4 1 6
..txa4, Chekhov-Murdzia, Polish Cht
(Lubniewice) 1 994, and now Kapen­
gut gives 1 6...lZJc7 leading to compli­
cations, whereas Chekhov's own line
1 6...lZJb4 17 i.bS ! a6 1 8 i.e2 "with
advantage" is contradicted by Schnei­
der: 1 8 ...�h4 ! 19 �h2 i.d4 20 i.d2
146 THE GAMBIT GUIDE TO THE MODERN BENONI
:e8 2 1 .ie 1 'ii'h6 22 .if2 .ig7 23
'ilVd2 lDf6 with at least equality.
b) 1 3 .ie3 l:Ib8 (after 1 3 ...lDc7 1 4
a4 a6 Kapengut gives I S lDc 1 bS 1 6
axbS .:tb8, and instead proposes I S
lDg3 .:tb8 1 6 eS ! ? dxeS 1 7 fS with com­
pensation) 14 a3 ( 14 'iid2 bS I S :ae 1
b4 1 6 lDd l lDc7 was unclear in the
game Chekhov-Stajcic, Kecskemet
1 99 1 ) 14 ... bS ! ? ( l 4 ....:te8 I S 'ii'd2 bS
1 6 b4 lDb6 is probably more accurate:
17 .td3 cxb4 1 8 axb4 lDxb4 1 9 .txbS
.td7 =) I S b4 lDb6 ! 1 6 .td3 cxb4 1 7
axb4 lDxb4 (Kapengut's 1 7 . . .lDa4
would be well met by 1 8 'ii'b 3) 1 8
.ixbS a6 1 9 .id3 fS ! 20 :b1 ! as 2 1
lDd4 fxe4 = (or 2 1 . . ..id7) Pau1sen­
Hartmann, Bundesliga 1 98617.
13 b3
Or:
a) 1 3 .id3 lDb6 ( 1 3 . . . lDaS ! ?) 1 4
lDg3 'ii'h4 ( 14 ...lDb4 I S .ib1 , Paul­
sen-Arnold, Bundesliga 1 985/6, is also
best answered by l S ...'ii'h4) I S 'ilVf3
lDb4 1 6 .ib1 .id4+ 1 7 'iil>h2 .id7 1 8
a3 lDa6 "with chances for both sides"
- Schneider.
b) 1 3 'iil>h 1 bS 1 4 b3 lDb6 I S .ib2
lDc7 = Arduman-Ghinda, Komotini
1 993.
13 lDa5!?
.••
The Perenyi theme again. 13 ... lDb6
has also been played with success, but
we have enough options already.
14 a3
1 4 .ie3 bS ! I S eS ! ? dxeS 1 6 lDxbS
( 1 6 fS? ! b4 ! 1 7 fxg6 { 1 7 lDa4 e4 }
1 7 . . . hxg6 1 8 lDe4 fS 1 9 lDgS f4 =+=
Shabalov) 1 6... exf4 17 .ixf4 'ilVb6 ! 1 8
lDbc3, and now 1 8 . . .c4+ with complex
play, Vilela-Andres, Havana 1 987; or
1 8 ...lDb4 intending 1 9 .ib 1 .ia6 !, as in
the game Nenashev-Shabalov, Tash­
kent 1 987.
14 b5 15 :bl b4 16 lDa2 bxa3 17
.•.
.ixa3 :b8 18 'iil>h2 .id7
The game is equal, Serper-Nena­
shev, Novosibirsk 1 989. It's significant
that Nenashev, having such trouble as
White in the last note, switched to the
black side of this line. In a few more
moves, he had an overwhelming ad­
vantage but (probably in terrible time­
trouble) managed to lose.
I would love to see more of this
dashing, provocative variation.
It is hard not to admire White's un­
pretentious build-up for a kingside
attack in the .id3 and lDge2 line.
Caught unawares, Black could surely
fall victim to the thematic pawn sacri­
fice ' 1 ' eS dxeS '2' fS, as even the
mighty Tal did versus Penrose (note to
Black's 1 6th move in Line A I ). But in
that particular variation, Black has a
wonderfully resilient position and
counterattacking chances that at least
equal White's attacking ones. Then, in
Line B , we saw Black engage in a sort
of absurdist Modernism. 9 ...lDg4 and
the journey of the knight to as via eS
and c4 may look outrageously stupid
to the Classical mind, but it has a mad
logic to it and as far as I can see, gives
Black unbalanced and legitimate play.
8 Ka pengut's 7 f3 System
1 d4 CDf6 2 c4 c5 3 d5 e6 4 CDc3 exd5 5
cxd5 d6 6 e4 g6 7 f3 i.. g7 (D)
w dian Defence ( 1 d4 CDf6 2 c4 g6 3 ttJc3
This set-up is called the 'Kapengut
System' (after its leading analyst, who
devoted 8 1 densely-packed pages to it
in his A65 monograph) or (as Kap­
engut himself called it) the 'Half­
Samisch Variation' (for reasons which
will become clear). With 7 f3, White
strengthens his centre and thus makes
it easier to restrain ... b5 (remember
that ... b5 tricks based upon ...CDxe4 are
common in the Benoni) White also
prepares i..e3 or i..g5-e3 without hav­
ing to worry about . . .CDg4. On the neg­
ative side, 7 f3 is rather slow. It
prevents White's knight from going to
f3, makes a future f4 a loss of tempo,
and somewhat weakens White's dark
squares.
I considered skipping this chapter
entirely (with coverage of 8 i..d3 else­
where) because the variations which
follow almost all arise from the
Samisch Variation of the King's In­
i..g7 4 e4 d6 5 f3) - that is, the lines of
that defence in which Black has
played . . . c5, White has answered with
d5, and Black has then played ...e6 fol­
lowed by ...exd5. The contemporary
reality is that the variations in this
chapter stem much more often from
the King's Indian move-order than
from the Benoni move-order, espe­
cially in the case of 8 i..e 3 and 8 i..g5 .
Gallagher's The Siimisch King's In­
dian; for example, covers all the lines
below except for 8 .td3.
Why then are the 7 f3 lines so often
listed under the Benoni? Because his­
torically, at the time ECO codes were
developed, White rarely played 6 Ji.g5
in the KID Samisch, and after 5 f3 0-0
6 Ji.e3, Black almost never played
6. . . c5 because it lost a pawn. It wasn't
suspected then that the latter sacrifice
might turn out to give Black remark­
able compensation and go on to be­
come the leading anti-Samisch move !
Since that sacrifice achieved popular­
ity, White has often switched to play­
ing 6. . .c5 7 d5, after which 7 ... e6 8
�d2 (or 8 ttJge2) 8 . . . exd5 9 cxd5
148 THE GAMBIT GUIDE TO THE MODERN BENONI
transposes into the subject-matter of
this chapter.
In addition, fewer players today em­
ploy 7 f3 against the Benoni because
other lines have become more popular.
Nevertheless, ECO itself, Informator,
and the Informator monographs are
based upon this opening code, and
other publications employ it as well.
For example, Psakhis organizes his
whole book around ECO codes, and
NCO follows the order of ECO with
minor and localized changes.
Since the Benoni player does have
to confront this variation, my compro­
mise has been to suggest thorough so­
lutions for Black but using (with one
exception) less complicated lines that
largely bypass the extremely dense
main lines. As a matter of priority, I
would rather devote precious space to
uniquely Benoni variations. Finally,
while I have been heavily dependent
upon Kapengut for this chapter's ma­
terial, you will still find a good deal of
original analysis, especially in the
main lines.
Without further ado, then, here are
White's moves:
A: 8 �d3 1 48
B: 8 lDge2 1 50
c: 8 �e3 1 52
D: 8 �g5 1 55
8 �b5+?! is rather illogical, since
White loses time with his bishop with­
out any corresponding gain: 8 . . .liJbd7
(or 8 ... �d7 =) 9 �f4 'iYe7 1 0 'ilVd2 a6
1 1 �e2 b5 ! ? (after 1 1 . . .0-0! White's
bishop is potentially misplaced on e2;
then 12 lDh3 b5 bypasses the �xb5
sacrifice in the next note) 1 2 lDh3 ( 1 2
�xb5 ! ? axb5 1 3 lDxb5 i s unclear after
1 3 ...lDxe4 14 fxe4 'ilVxe4+ and ...0-0)
12 . . .0-0 1 3 ..th6? (you can tell that this
is an old game !) 13 . . .lDxe4 ! 14 lDxe4
'iYh4+ 1 5 g3 'ilVxh6 1 6 'ilVxh6 �xh6 1 7
lDxd6 lDb6 + 1 8 lDxc8 �axc8 1 9 f4
IHe8 20 lDg 1 lDxd5 with an extra
pawn, Chekhover-Alatortsev, Mos­
cow 1 935.
A)
8 �d3 0-0 9 lDge2 (D)
B
This pOSItion resembles the last
chapter, but White has already played
f3. That move is unnecessarily com­
mittal and poses few problems. Be­
cause a few key 8 �d3 lines can
transpose to various instances of f3 in
the previous chapter, both Psakhis and
Schneider don't analyse 8 �d3 in this
move-order at all.
9...a6
9 . . .lDa6 is also fine. As this is a
fairly minor line, I won't give two so­
lutions.
10 a4
 KAPENGUT'S 7
This will be played soon anyway.
a) Mysteriously, the few sources
that even cover 8 ..id3 give 1 0 0-0, but
do not mention l O...bS in reply. Then
1 1 �h 1 or 1 1 a3 would be more to the
point than 1 1 �el tDbd7 1 2 lWh4
�b6 ! ? 1 3 ..ie3 tDeS 14 ..ic2 M I S
tDd l as 1 6 b 3 ..ia6 =1= Takemoto-Vag­
anian, Teesside U-26 Wcht 1 974. No­
tice also that 1 1 a4? ! c4 1 2 ..ibl M 1 3
tDa2 �b6+ 1 4 �hl as favours Black,
who has the moves . . ...ia6 and . . .tDbd7
in store. In any case, l O...bS should be
at least adequate in response to 10 0-0.
b) 1 0 ..ie3 ( , ? ! ' Kapengut, but that
goes too far) 1 0 . . .bS 1 1 a3 l::te8 1 2
�d2 tDbd7 1 3 M (the point of 1 1 a3;
not 1 3 i.h6? tDxe4 !) 1 3 . . . cxb4 (I
would prefer the elastic 1 3 ... ..ib7 1 4
0-0 .uc8 I S :tac 1 t2Jes = targeting c4
and d3) 14 axM i.b7 I S 0-0 l::tc8 1 6
nac 1 ( 1 6 tDd4 l::txc3 ! 1 7 'iVxc3 tDxdS
1 8 exdS { I 8 'iWd2 tDxe3 1 9 �xe3 'iWb6
+ } 1 8 . . Jhe3 1 9 �f2 tDb6 ! ? + Kapen­
gut; excellent analysis, although the
final position is resignable; e.g., 20
'uad l 'iWh4+) 1 6 . . .tDb6 (I don't like
this at all; 1 6 . . .tDeS 1 7 tDd4 tDxd3 1 8
�xd3 tDd7 { or 1 8 . . .:c4 } must be
better) 1 7 tDbl ! tDfd7 1 8 lhc8 ..ixc8
19 tDd4 i.b7 and now Kapengut likes
Black after 20 tDb3 tDeS 2 1 tDaS ..ia8
22 .l:tc l fS !, Guigones-Demarre, French
Cht 1 989, but simply 20 l::tc I ! confers
a large advantage on White (20... tDeS
2 1 tDa3 !).
10... tDbd7 1 1 0-0
Delaying or omitting this move is
ineffective:
a) 1 1 h4 hS 1 2 ..igS tDeS 1 3 ..ic2
.td7 14 �c l ? ! bS IS axbS axbS 1 6
f3 SYSTEM 149
l::txa8 �xa8 gives Black a slight advan­
tage, Merlini-Derieux, Cannes 1 995,
in view of 17 'iWf4 ! (best) 17 ... b4! 1 8
i.xf6 ( 1 8 tDd l i.a4 ! having in mind
19 i.xa4?? tDd3+) 1 8 . . . �a l + 19 tDd l
b3 ! =1=.
b) 1 1 i.f4 'ilie7 1 2 �d2 tDeS 1 3
i.xeS �xeS offers Black at least
equality (two bishops, dark squares),
1.Andreasen-Ruxton, Amhem jr Ech
1989/90.
11 tDe5
...
Any logical move equalizes here;
e.g., 1 l ...�c7, 1 l .. .l::tb8 or 1 l . . .l::te 8.
12 i.e2
Or:
a) 1 2 f4? ! is premature: 1 2 ...tDxd3
1 3 'iWxd3 l::te8 14 tDg3 ..id7 + (or
1 4. . .tDg4 +).
b) 12 i.e3 l::te 8 13 b3 ( 1 3 'fid2 :b8
14 'ufbl tDxd3 I S 'ilixd3 tDd7 16 M fS !
=1= Renet-Haik, Paris 1 986) 1 3 . . . tDxd3
(or 1 3 . . . :b8) 14 'tWxd3 l::tb 8 = (or
14 . . .tDd7 =) .
12 .l:tb8 13 f4
•.
1 3 l::tb l prepares M, but 1 3 . . .bS 1 4
axbS axbS I S b4?! fails to hold the bal­
ance after I S ...cxb4 16 l::txb4 �aS 1 7
l::tb3 M 1 8 tD b l ..ia6 =1=.
13 tDe4
..•
1 3 . . .tDed7 is also playable, since
White has wasted a move with f3-f4
and is faced with . . .bS ; e.g., 14 tDg3 bS
I S axbS axbS 16 eS dxeS 1 7 fS ..ib7
( 1 7 ...e4 ! ?) 1 8 d6 and apart from 1 8...M
1 9 tDce4 ..ixe4 with an edge for Black,
Ravikumar-Murshed, Calcutta 1 986,
Kapengut suggests 1 8 ...'ilib6.
After the text-move ( 1 3 . . . tDc4),
Lida Garcia-Niegovich, Olivos 1 993
went 1 4 b3 tDb6 ( 1 4 ... tDaS ! ?) I S as
150 THE GAMBIT GUIDE TO THE MODERN BENONI
ttJa8 ! 1 6 l:.a2 ttJc7 (a manoeuvre that
comes up in both the Benoni and
Benko Gambit) 1 7 i.d3 :e8 1 8 ttJg3
bS =.
B)
8 ttJge2 (D)
B "iYb6 1 9 i.e2 :a2 with compensation,
This has become more common in
the last few years. It can very easily
transpose to Line C (8 i.e3) or D (8
i.gS). We will look at an independent
line of particular interest.
8 0-0 9 ttJg3
•••
9 i.f4 ttJhS 1 0 i.e3 fS leaves Black
very active; an attractive possibility
would be 1 1 exfS :xfS ! 1 2 "iVd2 ttJd7 !
with the point that 1 3 g4 is answered
by 1 3 ...ttJeS ! 1 4 ttJgl ( 1 4 i.g2 ttJc4)
1 4...l:.xf3 I S gxhS i.h6 ! ! .
9 ttJhS!?
...
A new, eccentric, and fascinating
idea in this position. The same idea
can also be played with ... a6 and a4
thrown in, but it may be better to pre­
serve the idea of ...ttJa6.
Since 9 ...ttJhS is so speculative, I
should point out that older techniques
such as 9 ... a6 1 0 a4 ttJbd7 still work
here. But I think the most appropriate
alternative is 9 ... ttJa6, if only because
the usual problem with . . .ttJa6 -
White's attack by f4 - comes at the
cost of a critical tempo and tends to be
too slow. Briefly: 9 ... ttJa6 10 i.e2
ttJc7 1 1 0-0 l:.b8 ( l l .. .a6 12 a4 ttJd7 1 3
i.f4 ttJeS 1 4 lIb l hS I S 'iWd2 bS ! ? 1 6
axbS ttJxbS 1 7 ttJxbS axbS 1 8 i.xb5
Novikov-Gufeld, Tbilisi 1 988) 1 2 i.g5
( 1 2 i.f4 bS 1 3 'iVd2 :e8 14 'lith l was
Christiansen-Nunn, Munich 1 99 1 ;
Black can then play 1 4...b4 with ideas
like . . . a5 and ... ttJbS-d4) 1 2 ... h6 1 3
i.e3 l:te8 1 4 "iVd2 'lith7 1 5 a4 a6 1 6
f4 ! ? bS 1 7 eS b4 1 8 ttJce4 ttJfxd5 1 9
l:tad l dxeS 2 0 fS i.b7 2 1 i.xc5 ttJf4!
with an attack, Czerwonski-Kamin­
ski, Polish Ch 1 994.
10 ttJxhS gxhS (D)
w
Now Black has permitted his h­
pawns to become doubled, and unlike
other comparable variations, he hasn't
even eliminated White's light-squared
bishop in the process (i.e., by i.e2 and
 [(APENGUT'S 7
i.xhS). But the pawn being on f3 weak­
ens White' s dark squares, he is behind
in development, and .. .fS promises to
cause trouble on the kingside.
11 i.d3
This move has been played in al­
most all games with 9 ... lbhS thus far.
Otherwise:
a) 1 1 i.e3 fS 12 'iVd2 would also be
logical, when a sample continuation
might be 1 2 .. .fxe4 ( 1 2 . . . a6 1 3 a4 fxe4
is similar) 1 3 lbxe4 i.fS 14 i.d3 lbd7
( l 4 . . . c4 ! ?) I S 0-0 ( 1 S lbgS 'iVb6 1 6
�bI .l:.ae8 with good play; e.g., 1 7
lbe6 i.xe6 1 8 dxe6 lbeS !) I S . . .lbeS
16 ..te2 i.xe4 (or 1 6 ...'iVd7 1 7 .l:.ael
i.xe4 18 fxe4 lbg4 =) 17 fxe4 lbg4 (or,
again, 1 7 ...'iVd7) 1 8 l:hf8+ 'iVxf8 1 9
l:tfl 'iVe7 20 i.xg4 hxg4 2 1 l:tf4 h S =.
b) 1 1 i.f4 was played as I was
writing this chapter (I suspect the
theory will be much more fleshed out
by the time you read this !): l 1 . ..fS 1 2
'iVd2 'iVf6 1 3 i.gS 'iVg6 ( 1 3 . . .'iVeS ! ? is
also possible, to answer 14 i.d3 with
14 ...lbd7 I S 0-0 f4 1 6 i.h4 a6 1 7 a4
and 1 7 ...':'b8, or even 1 7 . . .'iVe8 intend­
ing ... lbeS) 14 i.d3. Here, instead of
the committal 14 ...lba6, as in Ward­
G.Buckley, British Ch (Millfield) 2000,
this appears to be the right time for
14 . . .fxe4 ! I S lbxe4 ( 1 S i.xe4 i.fS; I S
fxe4 ..td4 1 6 i.f4 lbd7 =) I S . . ...tfS 1 6
0-0 lbd7 = with active piece-play.
11 f5 12 0-0 (D)
•..
1 2 exfS i.xfS 1 3 0-0 ( 1 3 i.xfS
l:xfS 14 0-0 lba6 followed by . . . lbc7,
targeting the d-pawn - Mortensen)
13 ... lba6 ( 1 3. .. i.xd3 ! ? 1 4 'iVxd3 lbd7
"also looks OK" - Mortensen) 1 4
i.xa6 ! ? bxa6 I S ..te3 1:.b8 ! ? (perhaps
f3 SYSTEM 151
more accurate is Ward's suggestion of
I S ...1:.e8 or Mortensen' s of I S . . .'iWh4)
1 6 'iVd2 'iVf6 1 7 l:tac l 'ili'g6 1 8 b3 h4! ?
1 9 lbe2 i.d3 2 0 lbf4 ':'xf4 2 1 i.xf4
i.xfl 22 �xfl (22 ':xfl c4! followed
by ... c3 - Mortensen) 22 ...h3 23 g3
1:.f8 24 .l:.e l (24 'iVe2 1:.e8 2S 'iWxa6
i.d4! intending ...'ii'hs , and on 'iVd3,
.....te3) 24.....teS (24...i.d4! ?) 2S ..txeS
lhf3+ 26 'itg l (26 �e2? 'iig4 is too
strong) 26 ...dxeS 27 ':xeS ':xg3+ 28
hxg3 1Wxg3+ 29 �h l 'iVxeS 30 d6
'iVe4+ = Ward-Mortensen, Copenha­
gen 1 999.
12 lba6
•.•
An all-purpose move, clearing the
back rank, threatening . . .c4 in many
positions and preparing . . .lbc7 and
...bS while staying in touch with fS.
Nevertheless, a different and promis­
ing move is 1 2... lbd7:
a) 1 3 exfS lbeS =.
b) 1 3 lbe2 ! ? c4 and now 14 i.c2
1/2-1/2 was the finish of Dreev-Bologan,
Beijing 2000. Instead, 14 ..txc4 'iWb6+
I S 'Oth l lbeS gives Black excellent
compensation.
152 THE GAMBIT GUIDE TO THE MODERN BENONI
c) 1 3 SLc2 lDeS 1 4 lDe2 ' ! ' 'iih4' ! '
(Dreev's annotations, but I like the
look of 14 ... lDg6, practically forcing
I S exfS SLxfS 1 6 SLxfS lhfS and if 1 7
lDg3, then 1 7 .. .l::tf8 ! 1 8 lDxhS? ! SLd4+
1 9 �h l 'iih4 and now 20 lDg3? ! SLeS
or 20 g4 'iWh3 intending .. Jlae8, when
White is very tied down) I S f4 ! lDc4
( l s ...lDg6) 1 6 lDg3 lDxb2 1 7 'iWe2 'iWg4
1 8 .l:.f3, Dreev-Peng Xiaomin, Beijing
2000. At this point Fritz suggests the
remarkable 1 8 ...lDdl ! !, when the ob­
vious line would seem to be 1 9 'iWxd l
SLxal 20 lDxfS �xfS 2 1 :g3 .i.d4+ 22
�f1 'iWxg3 23 hxg3 SLg4. This may
still be unclear, but I'd rather be Black.
13 a3
Or:
a) 1 3 .i.f4 ! ? fxe4 14 fxe4 c4 ! I S
SLxc4 'iWb6+ 1 6 �h l 'iWxb2 1 7 SLxd6!
and now Black chose 17 . . .SLg4 ! ? in
Dreev-Schekachev, Russian Ch (St
Petersburg) 1 998, when Black won a
piece but White's pawns were strong
and there was an early, justifiable draw.
Similar play would have resulted from
1 7 ...:xf1 +.
b) 13 SLe3 is again untried: 13 ...:b8
1 4 'ii'd2 lDc7 I S a4 a6 looks solid;
Black could also try 1 3 . . .f4 ! ? 14 .i.f2
.i.eS and attempt to play on both sides
of the board.
13 SLd7 14 'ii'c2 'ii'f6
•.•
It's difficult to decide whether to
play ...f4 in this line. One possibility is
1 4 ... f4 ! ? IS lDe2 SLeS 16 .l:.bl bS 1 7
b 3 lDc7 1 8 .i.d2 'iWgS 1 9 SLaS lDe8 20
�h l lDf6, thinking about ... h4 and
. . . lDhS. Nevertheless, I prefer the
text-move.
15 �hl l:tac8
A different possibility is I S ... :ab8
1 6 �e3 bS ! ?
16 SLe3 c4 17 SLe2 b5!? 18 f4
An intended improvement over 1 8
SLxa7 lDcS 1 9 .l:.ad l , when, instead of
19 ...�h8 20 SLXCS l:txcs 21 f4 ;j; Cher­
nin-Glek, 2nd Bundesliga 1 998/9,
Glek suggests 19 ... f4 ! with compensa­
tion. I agree, since White has a hard
time finding anything constructive
while Black improves his position (20 .
l:tgl M).
18...lDc5 19 e5!?
19 �xcS :xcS 20 .i.xhS is critical,
pitting the two bishops against an ex­
tra pawn; e.g., 20. . :�M (20... aS) 21
�e2 as and Black has active play, but
it's hard to assess this.
After the text-move ( 1 9 eS ! ?),
Khenkin-David, French Cht 1 998/9
went 1 9 ...dxeS 20 fxeS 'ilVxeS 2 1 SLf4
'ilVf6 22 d6 (22 SLxhS lDd3 +) 22 ...�h8
23 :ad1 112_ 112. Black has ideas like
... as and ...b4, whereas White has a
well-supported passed pawn.
I have to admit that 9 ...lDhS ! ? is aw­
fully committal, and some players
may find Black's pawn-structure ob­
jectionable. Nevertheless, Black en­
joys active play, and sometimes it's
fun to bypass standard theory. Alter­
natively, for the cautious at heart, I
would recommend 9 ...lDa6 as a sound
alternative that doesn't require a great
deal of study.
C)
8 SLe3 0-0 (D)
Once White makes his next move,
we have transposed into a 'pure'
Sfunisch King's Indian, Le., 1 d4 lDf6
 [(APENGUT'S 7
2 c4 g6 3 liJc3 !iLg7 4 e4 d6 5 f3 0-0 6
.te3 c5 7 d5 e6 [8 'ii'd2 or 8 liJge2]
8 ...exd5 9 cxd5 . We examine:
Cl: 9 liJge2 1 53
C2: 9 'ii'd2 1 54
C1)
9 liJge2
Now instead of transpositions to
fashionable lines involving the move
...h5, there is an intriguing alternative:
9 .a6 10 a4
..
Black experiences few difficulties
if White allows ...b5 : 1 0 liJg3 b5 1 1
.te2 liJbd7 1 2 0-0 liJe5 1 3 'ii'd2 :e8
14 b3 ( 14 !iLh6 !iLh8 1 5 liJdl liJfd7 1 6
:c1 liJb6 1 7 b 3 f6 1 8 i.e3 !iLd7 1 9 h4
Cjjf7 20 h5 f5 ! with double-edged play,
Ward-Gallagher, British Ch (Scar­
borough) 1 999) 1 4 ...i.d7 1 5 a4 bxa4,
Begovac-Wojtkiewicz, Bern 1 993, and
now 1 6 bxa4 ! 'ii'a5 is best, with equal­
ity.
10... liJbd7 11 liJg3 liJe5 12 !iLe2
i.d7
13 f4!?
A confrontational move, but com­
plications will follow in any case:
f3 SYSTEM 153
a) Psakhis's proposal 1 3 h3 is met
by 1 3 . . .b5 ! ( 1 3 ...liJh5 ! ?) 1 4 f4 ( 1 4
axb5 axb5 1 5 liJxb5 !iLxb5 1 6 !iLxb5
'ii'b6 1 7 !iLe2 'ii'xb2 with equal play)
1 4 ...liJc4 l 5 !iLxc4 bxc4 1 6 0-0 l:tb8 1 7
'ii'd2 l:tb3 1 8 f5 'ii'e7 1 9 l:tf3 l:tfb8
gives Black a slight advantage, Nena­
shev-Banikas, Khania 1 999.
b) 1 3 0-0 b5 ! 14 h3 ( 1 4 axb5 axb5
1 5 !iLxb5 i.xb5 1 6 liJxb5 'ii'b6 + Seir­
awan-Ivanchuk, Roquebrune blitz
1 992) l4 ...l:tb8 1 5 b3 liJe8 ! 1 6 l:te l
'ii'h4 with plenty of play, Gallagher­
Nunn, London Lloyds Bank 1 990.
13 liJfg4!
•••
This entertaining approach, sacri­
ficing a piece, was the product of over­
the-board inspiration by the Bristol
player Chris Beaumont. For those who
don't believe in this idea, 1 3 ...liJeg4
1 4 !iLd2 h5 1 5 h3 liJh6 is still held to
be equal. But, like 1 3 ... liJfg4, it has
barely been tested.
14 !iLgl! 'ii'h4 15 fxe5
Equality results from 1 5 'ii'b3 b5 1 6
fxe5 !iLxe5 (or 1 6. . .c4) 1 7 liJd 1 bxa4
1 8 'ii'f3 f5 =.
15 !iLxe5 (D)
••.
w
154 THE GAMBIT GUIDE TO THE MODERN BENONI
From the diagram, analysis by Lev­
itt goes 1 6 'Yi'd3 ! ( 1 6 i.xg4? i.xg3+
1 7 hxg3 'ili'xh 1 1 8 c;¥;>f1 fS ! was good
for Black in Levitt-Beaumont, British
League (4NCL) 1 995/6) 1 6 . . . c4 ! 1 7
'Yi'f3 fS 1 8 0-0-0 ( 1 8 exfS .txfS +)
1 8 . . . fxe4 1 9 "iWxe4. Now instead of
Levitt's 19 ...'iWgS+, against which Nunn
offers 20 :td2! i.f4 2 1 'ikd4 ! ±, I pre­
fer 1 9 ...:tae8 !, and without pretending
to know what's going on, I would as­
sess this as dynamically balanced. A
sample line derived from Nimzo 7.32
and Fritz 6 (with some disciplinary
guidance) goes 20 li'xc4 (20 .txg4
i.xg4 is too strong) 20...bS ! ? 2 1 'Yi'c7
(21 "iWb4 ! ? as 22 'ili'xaS i.xg3 23
i.xbS i.fS ! is unclear) 2 l . ...l:.e7 ! 22
i.d4 .tf4+ 23 c;¥;>bl liJxh2, and so
forth, ending in equality. Entertaining
stuff!
C2)
9 'ikd2 liJa6!?
A rare approach, but one with a
good reputation. As explained above,
...liJa6 is particularly appropriate when
White has played f3 because it would
take two moves for him to play the
move Black most fears, i.e., f4. As al­
ways, 9 ... a6 1 0 a4 l:.e8 (or 1 O ... liJbd7)
is quite playable.
10 ttJge2
1 0 i.d3 could lead to some curious
play after 1 O ... l:.e8 1 1 ttJge2 i.d7 ! ?
( l l ...ttJd7 1 2 f4 ttJM 1 3 i.bS { 1 3 i.bl
ttJb6 ! } 1 3 ... a6 ! ! 1 4 i.xd7 i.xd7 IS a3
as ! was the amazing course of Pliasu­
nov-Simantsev, St Petersburg Chigorin
mem 2000; Black stands well!) 1 2
ttJbS ! ? iVb6 1 3 a4 ttJM ! 1 4 li'xM ( 1 4
i.c4 a6 I S as?? axbS) 14 ... a6! I S 'Yi'b3
( 1 S 0-0 ttJxdS !) I S . . . axbS 1 6 0-0 'ikaS !
= with the idea 1 7 i.xbS i.xbS 1 8
�xbS ttJxdS.
10 ttJc7 (D)
••.
1O ... i.d7 ! ? deserves consideration:
1 1 g4 ! ? ( 1 1 ttJg3 lIe8) l l .. .bS 1 2 ttJg3
M Kapengut; then 1 3 i.xa6 bxc3 1 4
bxc3 �aS threatens ... ttJg4.
11 ttJg3
Or 1 1 ttJc 1 ltb8 1 2 a4 ( 1 2 eS ttJfe8!
13 exd6 ttJxd6 1 4 .txcS l:.e8+ IS .te2
ttJc4) 1 2 . . . a6 1 3 as :te8 1 4 i.e2 bS 1 5
axb6 l:.xb6 =.
11..J�b8?!
1 1 . .J�e8 ! is more accurate, not al­
lowing eS : after 12 .te2 a6 1 3 a4 l':.b8
14 as i.d7 IS 0-0 ttJbS the position is
equal.
12 a4
Playing 1 1 .. J::te8 first would have
avoided 1 2 eS !, which I like for White ·
after 1 2 . . . ttJfe8 1 3 exd6 ttJxd6 1 4
O-O-O! ? b 6 I S ttJce4 ! ttJce8 1 6 i.e2 ;!;.
12...l:.e8 13 i.e2 a6 14 0-0 b5
Damljanovic-Ivanovic, Yugoslav Ch
(Kladovo) 1 990 continued I S axbS
 KAPENGUT'S 7 f3 SYSTEM
axbS ( l S ... lLJxb5 ! ?) 1 6 i.h6 b4 1 7 l2Jdl
i.d7 1 8 i.xg7 �xg7 19 lLJe3 lLJbS 20
i.xbS i.xbS 2 1 l::tf2 ( ,unclear' - Kap­
engut) 2 l . ..�g 8 ! 22 b3 l::ta8 23 l::txa8
"i¥xa8 24 iib2 �d8 and ... lLJd7 =.
D)
8 i.g5 0-0 (D)
w �e7 = 17 .:tfel 1He8 1 8 i.n l::tac8 1 9
9 'ilVd2
After 9 lLJge2, play tends to con­
tinue 9 ... a6 1 0 a4 lLJbd7 1 1 lLJg3 hS 1 2
i.e2 �aS, when 1 3 �d2 .:te8 trans­
poses to the main line. If 1 3 0-0, Black
has 1 3 ... lLJh7 ! = with the idea that 1 4
i.e7? is met b y 1 4...:e8 I S i.xd6 h4
16 lLJh l �b6 =t.
9.. .a6
Those looking for something effec­
tive but out of the ordinary should
check the theory on 9 ...i.d7.
10 a4
The idea of allowing . . . bS had a
brief spurt of interest 4-S years ago,
but has again faded into obscurity. An
abbreviated overview of typical play:
10 lLJge2 lLJbd7 1 1 lLJg3 bS 1 2 i.e2
and then:
155
a) 12 . . .c4 (this has the advantage
of reorganizing the queenside without
delay) 1 3 0-0 lLJcS 14 'iil'h l i.d7 (or
14 .. J�e8 IS �f4!? �e7 1 6 �h4?, Lau­
tier-Xie Jun, Monte Carlo rpd 1 996,
16 . . . h6! 17 i.xh6 lLJfxe4! =t) IS i.h6
(the point of Black's move-order is to
hold his own on the queenside after,
for example, I S lbbl as ! 1 6 a4 b4 1 7
lLJb5 lLJxa4 1 8 i.xc4 �b6 1 9 i.e3
lLJcs 20 lLJd4 a4) I S ... i.xh6 1 6 �xh6
.:tac 1 as with equal chances, Dreev­
Van Wely, London ECC 1 996.
b) More conservative is 1 2 . . .:e8
1 3 0-0 l:tb8 ( 1 3 ...c4 is similar to line
'a') 14 a4 ( 1 4 .:tabl �aS = ) 14 ...c4 I S
axbS axbS 1 6 l:ta7 lLJcS 1 7 i.e3 b4 1 8
lLJa4 ( 1 8 lLJbl 'ilVb6) 1 8 ... c3 1 9 bxc3
lLJxa4 20 l:txa4 bxc3 =.
10 l:te8 11 lLJge2
.•.
1 1 i.e2 hS 1 2 i.dl ? ! is too slow:
1 2 . . .�aS 1 3 l::ta3 (to prevent ... b5)
1 3 ...lLJbd7 14 lLJge2 lLJe5 I S b3 b5 and
Black has a clear advantage, Tegsh­
suren-Shulman, Sioux Falls 2000.
1l ...lLJbd7 12 lLJg3
Or:
a) A popular option is 1 2 lLJd l ,
when one important line goes 1 2...lLJe5
1 3 lLJec3 �aS 14 i.e2 ( 14 ':'a3 �4 !)
1 4 ...b5 1 5 0-0 lLJfd7 1 6 lLJf2 lLJc4 17
i.xc4 ( 17 axb5 �xal 18 .:txal lLJxd2
1 9 i.xd2 l:tb8 ! ?) 1 7 ...bxc4 1 8 f4. This
is the often-quoted Meulders-Douven,
Tilburg 1 993, which went 1 8 .. Jlb8 1 9
eS dxe5 20 lLJfe4 ! �b6 2 1 f5 ! intend­
ing 2 l .. .�xb2 22 �el ! with a terrific
attack. But simply 1 8 . . .h6 1 9 i.h4
.l:tb8 looks much better, since 20 e5
dxe5 2 1 lLJfe4 'ii'b6 22 fS is now
156 THE GAMBIT GUIDE TO THE MODERN BENONI
unconvincing due to 22 ... gS 23 i.f2
'Wxb2. I also think that 1 8 ... i.d4 ! mer-
its strong consideration. In my opin­
ion, this line has been overrated.
b) 1 2 liJc 1 h6 ! ? ( 1 2 ...:b8 1 3 i.e2
�c7 is another approach, from the
game M.Franco-Tal, Varna OL 1 962)
1 3 i.e3 ( 1 3 i.xh6? liJxe4 +) 13 ...hS 1 4
.ie2 liJeS I S 0-0 liJh7 1 6 :a3 ( 1 6
liJ1a2 fS ! ; 1 6 l'.:tb1 fS 1 7 b4 b6 1 8 �h 1
liJf6 1 9 exfS i.xfS =) 16 ...fS 1 7 liJ 1 a2
( 1 7 h3 liJf6 1 8 f4 liJf7 = with control
of e4) 1 7 ...liJf6 1 8 as fxe4 1 9 liJxe4
liJxe4 20 fxe4 i.g4 and Black has
what he wants, having isolated White's
e-pawn and secured eS, Alterman­
Gelfand, Riga 1 987.
12 hS
•••
Or 1 2 ...�a5 first. The move ... hS in
these positions anticipates a later
...liJh7 and/or ...h4.
13 .ie2 'ti'aS 14 0-0 c4!
Very aggressive. I think that the al­
ternative 14 . . .liJh7 ultimately equal­
izes, but it is harder going.
IS i.e3
I S �h l ? liJh7 ! ( 1 S . . .�b4 = ) 1 6
.ixc4 (not 1 6 i.e3?? h4, when Black
wins a piece !) 1 6 ...'Wb4 1 7 i.d3 liJxgS
1 8 'WxgS liJcS =t.
IS liJeS 16 liJhl
.••
Probably best, in order to reorga­
nize by liJf2.
16 liJfd7 17 f4?!
•.•
This exposes White's centre too
much. Better is 1 7 liJf2 liJcS ! =. The
more ambitious 1 7 liJb1 is also possi­
ble. Then following 1 7 ...'Wc7 1 8 liJa3,
Black can respond directly by 18 ... c3 ! ?
1 9 bxc3 liJcs = , hitting b 3 and a4, or
he can maintain equal chances by
1 8 ...ltJcS 19 1i'b4 ( 1 9 �c2 fS ! =) 19 .. .fS
(or 1 9 ... .id7 20 liJxc4 liJxc4 2 1 'Wxc4
.ixa4 ! = ) 20 liJxc4 liJxc4 and ...fxe4
with equality. Instructive play: the
sacrifice of the c-pawn for open lines
and play against weaknesses is ubiqui­
tous in the Benoni.
17 liJg4!
••.
The sharpest move. 1 7 ... liJd3 ! ? 1 8
i.xd3 cxd3 1 9 liJf2 liJcs 20 i.xcs
'WxcS 2 1 �h 1 i.d7 ! produced mutual
chances in Antonsen-Tseshkovsky,
Copenhagen 1 996.
18 i.d4 liJcS!
Here 1 8 ...i.xd4+ 19 'ii'xd4 'it'cs =
has been suggested, but the text-move
( 1 8 ...liJcS !) appears even stronger, as
1 9 .ixc4 ( 1 9 i.xg7 liJb3) 1 9...liJxe4 20
�d3 liJxc3 2 1 'ilkxc3 (21 i.xc3 �cS+
22 liJf2 i.fS) 2 l .. .�xc3 22 bxc3 liJe3
23 l'.:tfel i.xd4 24 cxd4 i.d7 yields a
substantial and perhaps winning posi­
tional advantage after ... liJfS.
While they require some serious
study, I don't think that these 'Half­
Samisch' lines should worry the
reader. Unlike systems in other chap­
ters (e.g., S and 6), White has few at­
tacking ideas, and positionally, Black
has as many effective strategies as
White. In addition, Black has resort to
more than one satisfactory line against
each main line so that he needn' t be re­
liant upon one tactically-dependent
solution.
9 M odern M a i n L i ne
1 d4 tDf6 2 c4 cS 3 dS e6 4 tDc3 exdS S
cxdS d6 6 tDf3
Fortunately, this move (tDf3) is al­
ready in for a great many Benoni play­
ers (who use the move-order 1 d4 tDf6
2 c4 e6 3 tDf3 c5), because the move­
order with 6 e4 g6 has some subtle
transpositional problems that aren' t
even mentioned in the books. To begin
with, 7 J..d3 doesn't always lead to the
main lines of Chapter 7 or to this chap­
ter; for example, 7 J..d3 J..g7 8 J..g5
and 7 J..d3 J..g7 8 h3 0-0 9 J..g5 are
unique and can be found at the begin­
ning of Chapter 7. In the last line, 9
tDf3 transposes to Line B2 of this
chapter.
It is remarkable how few players
have played 6 e4 g6 7 J..d3 with the in­
tent of entering the Modem Main Line
by 7 ... J..g7 8 h3. Even fewer players
have used 7 h3 ! ? for the same purpose.
After 7 h3 J.. g7, either 8 .i.d3 0-0 9
tDf3 or 8 tDf3 0-0 9 J..d3 again sends
us to Line B2. Although White hasn't
yet used 6 e4 g6 7 h3 for transposi­
tional purposes, he certainly could,
and one wonders if Black can do any­
thing to exploit this early h3. Perhaps
some development involving an early
... a6 and . . :�a5 could be investigated,
but for now, it appears that this is a
clever way for White to bypass Line A
of this chapter (assuming, once again,
that he isn't already committed to
tDf3).
6 g6
•••
Now we look at two ways for White
to try to head for the Modem Main
Line, which is characterized by the
moves tDf3, h3 and J..d3 :
A: 7 h3 1 57
B: 7 e4 1 6 1
A)
7 h3
This move aims for 7 ... J..g7 8 e4
0-0 9 J..d 3, which is Line B2 of this
chapter. Black can either agree to con­
test those lines, or circumvent them in
the following way:
7...a6
7 .. :�e7? fails to stop 8 e4 ! , due to
8 . . . tDxe4?? 9 'iVa4+.
8 a4
8 e4 b5 will be discussed in Line
B 12.
8...'iVe7 (D)
This is a pretty comfortable line for
Black. Once e4 is prevented, he can
develop without worrying about a cen­
tral breakthrough by his opponent.
9 J..gS
The most common move, but not
necessarily best. Alternatives:
a) 9 g3 is slow with h3 already on
the board: 9 . . . .i.g7 1 0 .i.g2 ( 1 0 tDd2 ! ?
;t was given by Ionescu, but 1 O...tDbd7 !
158 THE GAMBIT GUIDE TO THE MODERN BENONI
compares favourably with line 'd' )
1 0 ...0-0 (or 1 O...tLJe4 1 1 tLJxe4 'iVxe4
1 2 0-0 0-0 1 3 tLJd2 'iVe7 14 tLJc4 tLJd7
= Atalik-Ionescu, Mangalia 1 992) 1 1
0-0 tLJbd7 1 2 tLJd2 tLJh5 ! ? 1 3 cJth2 f5 !
14 f4 tLJdf6 1 5 tLJc4 i.d7 1 6 tLJb6 l:.tae8
= (control of e4) Arbakov-V.Milov,
Bad Ragaz 1 994.
b) 9 a5 tLJbd7 1 0 .ltg5 .ltg7 has no
real advantages for White over the
main lines after 9 .Jig5.
c) 9 l:.ta3 ! ? is a rare but interesting
approach. The idea is to play 1 0 e4,
meeting 1 O . . .tLJxe4?? by 1 1 tLJxe4
'iVxe4? 1 2 l:te3. The drawback to l:ta3
is that it is not a move White would
normally want in the e4/h3 system,
and is probably less useful than ...'iVe7.
Play continues 9 ... .ltg7 (an eccentric
idea would be 9 ...tLJbd7 1 0 e4 tLJe5 1 1
tLJxe5 'iVxe5 1 2 'iVf3 g5 ! ?) 1 0 e4 0-0 1 1
i.d3 and then:
c l ) l l ...tLJxe4 ! ? 12 tLJxe4 f5 is risky
but apparently playable:
c l l ) After 1 3 0-0 fxe4 14 l:.tel , in­
stead of 1 4... 'iVf7, as in Sinkovics-De
Sousa, Bagneux 1 996, a straightfor­
ward solution is 1 4 ...i.f5 1 5 g4 'iVf7
1 6 i.xe4 i.xe4 1 7 l:txe4 tLJd7 1 8 :f4
'iVe7 1 9 %:te3 tLJe5 =.
e l 2) 13 i.g5 'iVe8 ( 1 3 ... i.f6? 14
i.xf6 ':xf6 15 0-0 fxe4 1 6 l:.te 1 gives
Black serious problems) 14 0-0 fxe4
1 5 :el . Now Black should avoid
15 ... .i.xb2? 1 6 1he4 'iVf7 17 .lth6!, e.g.
17 ... .i.xa3 1 8 'iVaI .ltb4? ( 1 8 ....Jif5 1 9
l:.tf4 ± ) 1 9 tLJg5 +-, but 1 5 ... i.f5 ! 16
'iVbI ! 'iVf7 ! 1 7 .ltxe4 tLJd7 looks quite
satisfactory.
c2) l 1 . . .tLJbd7 (safer) 1 2 0-0 and
now, instead of 1 2 ... tLJe8, Labollita­
Rosito, Mar del Plata 2000, Black can
equalize more smoothly by 1 2 ...tLJh5 ! ;
e.g., 1 3 i.g5 i.f6 1 4 .i.e3 ( 1 4 .i.h6
l:.te8 ! 1 5 'iVd2 tLJe5 1 6 tLJxe5 .i.xe5 =)
1 4 ... tLJe5. Compare Line B2 where,
again, :a3 is not very useful.
d) 9 tLJd2 prepares tLJc4, to exploit
the weakness of b6. After 9 ...tLJbd7
(D), White has two main tries:
w
d 1 ) 1 0 tLJc4 tLJe5 I I tLJb6 :bS.
Black isn't upset about having his
problem bishop removed, and even
less so with a gain of time; e.g., 1 2
e4! ? ( 1 2 i.g5 h6 1 3 i.f4 .i.g7 1 4 e3
 MODERN MAIN LINE
i.f5 { l 4 . . . 0-0 1 5 ..te2 lLlfd7 = } 1 5
.te2 0-0 1 6 0-0 lLlfd7 17 lLlxd7 i.xd7
18 a5 f5 ! = Salgado Allaria-Ionescu,
Bucharest 1 993) 1 2. . . i.g7 1 3 .te2 0-0
14 0-0 lLled7 1 5 lLlxc8 l:lfxc8 1 6 i.f4
c4! 17 %te l b5 1 8 axb5 axb5 1 9 l:la6
lLle5 20 b4! cxb3 2 1 'ir'xb3 b4 22 lLlb5
tDxe4 23 'ir'xb4 lLlc3 24 �xd6, Shab­
tai-Lev, Tel-Aviv 1 992, and now it
looks as though 24 . . . lLlf3+ ! 25 �f1
(25 gxf3 lLlxe2+ 26 �f1 'iWxd6 +)
25 .. :�xd6 26 lLlxd6 lLlxe 1 27 lLlxc8
1:.b 1 is equal, since White must bail
out with 28 :'b6! l:la1 29 :a6 l:lb l .
d2) 1 0 e4 i.g7 1 1 .te2 0-0 1 2 0-0
is similar to a position from the Classi­
cal Main Line (see Chapter 1 0), but
White has played the premature h3.
Black has several satisfactory plans;
for example, 12 ...l:lb8 1 3 l:lel ( 1 3 a5
tDe8 14 !tel lLlc7 1 5 lLlc4 lLle5 !?; e.g.,
16 tDb6 lLla8 17 lLlxc8 :'fxc8 18 ..te3
tDc7 1 9 �d2 b5 20 axb6 l':!.xb6 with
double-edged play) 1 3 ... lLle8 14 i.f1
CiJc7 (a standard idea, to enforce . . .b5)
15 tDc4 ( 1 5 a5 b5 16 axb6 lLlxb6 =)
1 5 ... i.d4!? ( 1 5 ...lLle5 =) 1 6 i.e3 i.xe3
17 llxe3 lLle5 1 8 lLlb6 �f6 1 9 l:.b 1
CiJa8 20 lLlxa8 :'xa8 2 1 a5 i.d7 with
comfortable equality, Brito-Franco,
Asuncion 1 992.
e) 9 i.f4 (D) is probably better
than has been indicated by the lack of
mention in most sources.
Black can try several moves, but I
prefer these two:
e l ) 9 ...lLlh5 ! ? (an attractive move,
since the knight is well-placed here in
any case) 10 .tg5 ( 1 0 i.h2 i.g7 1 1
CiJd2 lLld7 1 2 lLlc4 lLle5 1 3 lLlb6 { 1 3
CiJxe5 i.xe5 14 i.xe5 �xe5 } 1 3 .. J:tb8
159
14 e4 f5; e.g., 1 5 i.e2 fxe4 ! 1 6 0-0 { 16
i.xh5 lLld3+ 1 7 �f1 O-O ! 1 8 i.f3
lLlxb2 + } 1 6 . . .lLlf6 with excellent
play) 10 ...f6 1 1 ..td2 f5 12 ..tg5 �f7 ! ?
(recommended b y Stohl, although
12 ...lLlf6 may be OK; for example, 1 3
lLld2 i.g7 1 4 lLlc4 0-0 1 5 e3 lLlbd7 1 6
i.e2 l:lb8 1 7 0-0 lLle5 =) 1 3 e4 h6 1 4
i.d2 i.g7 ! 1 5 exf5 i.xf5 = with the
idea 1 6 g4 i.xc3 1 7 i.xc3 i.e4 1 8
gxh5? i.xf3 +.
e2) 9 ... lLlbd7 and now:
e2 1 ) 10 lLld2 lLle5 can lead to a
fairly standard position after some­
thing like 1 1 e4 i.g7 1 2 i.e2 0-0 1 3
0-0 lLle8 intending " .f5 and/or ... i.d7
with . . .l:lb8 and . . . lLlc7, etc. Notice
that 1 1 i.xe5 'ir'xe5 1 2 lLlc4 'ir'e7 1 3
lLlb6 l:lb8, with . . . .tg7 and perhaps
...lLld7 to follow, encourages White to
exchange the problem bishop on c8, as
we have seen throughout this book.
e22) 10 e3 ..tg7 1 1 ..te2 0-0 1 2 0-0
lLle8 1 3 i.h2 ( 1 3 e4 l:lb8 14 l:le1 lLle5
15 lLlxe5 i.xe5 1 6 ..th6 lLlg7 17 f4
i.d4+ 1 8 �h2 i.d7 led to equality in
Greenfeld-Psakhis, Israeli Cht 1 997)
1 3 ...f5 14 lLld2 lLle5, and a sample line
160 THE GAMBIT GUIDE TO THE MODERN BENONI
is I S as t'iJc7 1 6 t'iJc4 t'iJxc4 1 7 .txc4
.td7 1 8 'iib3 .i.xc3 ! 1 9 bxc3 ( 1 9 l1Vxc3
.tbS = ) 1 9 ....tbS 20 lHb l lbb8 2 1
.txbS axbS 22 c 4 b4 2 3 .tf4 :a8 = .
9....t g7
9 . . .t'iJbd7 is similar, but in some
ways more flexible. The resulting play
is comparable to the main line, and
can complement its study:
a) 10 e3 h6 1 1 .th4 t'iJeS 12 t'iJd2
.tfS ! hits d3, and 1 3 e4 gS 14 .tg3 is
level after 14 ....tg6 (or 14 ... .th7, re­
serving the option of 1 6...t'iJg6 after I S
.i.e2 .tg7 1 6 0-0) I S .te2 .tg7 = ,
when one idea i s . . .hS-h4, or . . .t'iJfd7
intending . . .hS ; for example, 1 6 0-0
t'iJfd7 17 :e 1 hS 1 8 t'iJfl g4! with a se­
rious attack.
b) 1 0 t'iJd2 h6 1 1 .th4 .tg7 (D)
(here 1 1 ...t'iJes 12 e4 gS 1 3 .tg3 .tfS ! ?
resembles line 'a') and now:
w note to Black's 9th move (9 ... t'iJbd7) if
b l ) 1 2 t'iJc4 t'iJeS ! 1 3 t'iJb6 :b8
again challenges White to take the
problem bishop on c8, which he logi­
cally declines: 14 e3 0-0 (or 14 ... gS I S
.tg3 .tfS =) I S .te2 t'iJed7 1 6 t'iJc4
t'iJeS 1 7 t'iJb6 =. In 0stergaard-Emms,
Hillemd Politiken Cup 1 995, Black
tried to keep playing by 17 ...lIVc7 and
eventually won, although it is equal at
this stage.
b2) 1 2 e3 0-0 (again, I think that
1 2 ...t'iJeS intending ....tfS is worth a
look, since 1 3 f4 t'iJed7 14 e4 gS ! 1 5
fxgS t'iJh7 1 6 t'iJc4 t'iJxgS { or 1 6...hxg5 }
1 7 .td3 t'iJeS 1 8 t'iJxeS .txeS appears
fine) 1 3 .td3 :b8 14 0-0 t'iJeS I S .i.e2
gS 1 6 .tg3 t'iJe8 ( 1 6 ....i.fS ! ?) 1 7 f41
(this tends to leave kingside squares
weak even without the following tac­
tic) 1 7 ... t'iJd3 ! 1 8 .txd3 lIVxe3+ 1 9
�h2 lIVxd3 20 fxgS .i.xc3 2 1 bxc3
hxgS + Garza Marco-Suba, Saragossa
1 999.
10 e3
1 0 t'iJd2 t'iJbd7 1 1 e4 ! ? weakens
White's dark squares; e.g., 1 1 ...0-0 12
.te2 h6 1 3 .th4 gS 14 .tg3 t'iJeS in­
tending ... t'iJg6-f4.
10 h6
•.•
1 O . . .t'iJbd7 is comparable to the
White plays 1 1 t'iJd2; instead, 1 1 .i.e2
0-0 1 2 0-0 h6 1 3 .th4 t'iJeS 14 t'iJd2 g5
I S .tg3 .tfS gives Black perfectly
good play.
11 .th4 t'iJbd7
This move-order is preferred by John
Emms. Black wants to play ...t'iJeS and
....tfS . The immediate 1 1 ....tfS is also
reasonable, one example going 1 2
.td3 .txd3 1 3 �xd3 0-0 1 4 0-0 t'iJbd7
I S e4 :fe8 1 6 :fel gS 1 7 .tg3 t'iJh5"
1 8 .th2 t'iJeS 1 9 t'iJxeS .txeS with
equality, Komarov-Yudasin, St Peters�
burg 1 997 .
12 .te2 t'iJe5 13 t'iJd2 g5 14 .tg3
.tf5
 MODERN MAIN LINE
The key idea, preventing �c2 and
clearing the back rank to connect
rooks.
15 0-0
I S e4 .tg6 ( l S . . . .th7 ! ?, intending
.. .'�Jfd7 and . . . ltJg6 at some point, has
also been played) 1 6 0-0 0-0 17 .l:te 1
ltJfd7 1 8 ltJf1 fS (before ltJe3-fS be­
comes a problem) 1 9 exfS .txfS 20
ltJd2 �h8 2 1 as 'iWd8 22 ltJce4 �c7 23
b4 cxb4 24 'iWa4 ltJcS 2S 1ibc l and al­
though 2S .. :�d7 drew quickly in Lalic­
Ward, British Ch (Scarborough) 1 999,
2S . . .bS 26 �xb4 .l:.ac8 is safer.
15 ...0-0 16 a5 l:tae8!?
16 . . .ltJfd7 is probably more accu-
rate, since the best position for the
rooks is not yet known.
17 11el
17 lla3 can be met in the same man-
ner: 1 7 ... �c7 (to cover b7), and if 1 8
f4, 1 8 ... gxf4 1 9 .txf4 .th7 20 e4 ltJfd7
intending . . . ltJg6 is equal. The eS out­
post is extremely useful for strength­
ening Black's dark-square control,
and he may even be able to play .. .fS at
some point.
After the text-move ( 1 7 l:.e1 ),
17 . . .fIic7 1 8 e4 .th7 1 9 :a4 ltJfd7 =
was Garcia Ilundain-Spraggett, Candas
1992. Black has a solid defensive posi­
tion, and can play for . . .fS .
An attractive aspect of this 8 .. :�Ve7
line is that Black's problem bishop on
c8 tends to find useful employment on
the h7-bl diagonal. With correct play,
Black achieves full equality and posi­
tive chances .
B)
7 e4
161
This leads to our two repertoire
choices:
Bl: 7...a6 161
B2: 7....tg7 1 76
81)
7...a6 (D)
This is an attempt to bypass White's
h3/e4/.td3 system discussed in Line
B2. Either Black gets . . . bS in, or he
will be able to play 8 . . ..tg4 after 8 a4.
Several leading Benoni players have
embraced this system of development
in combination with the 7 h3 a6 8 a4
�e7 line that we examined in Line A .
Thus far, the results from this method
of play have been quite promising.
Whatever its theoretical consequences,
Black is certainly ensured of a dou­
ble-edged game thereby.
Notice that the immediate 7 . . ..tg4?
falls short after 8 'iVa4+ with the idea
8 . . .ltJbd7 (8 . . ..td7 9 'iWb3 costs Black
a tempo over standard lines) 9 ltJd2 ! ,
when White threatens f3.
After 7 . . . a6, White has two main
moves:
162 THE GAMBIT GUIDE TO THE MODERN BENONI
Bll: 8 a4 1 62
B12: 8 h3 171
Others:
a) We have already seen 8 �f4 b5
in Chapter 2.
b) After 8 �g5, 8 ... b5 seems rea­
sonable, or Black can play 8 . . . h6 9
�h4 g5 1 0 �g3 lLlh5, which trans­
poses to Line B of Chapter 3 and poses
no difficulties for him.
c) 8 �e2 b5 already threatens ...b4
and ... lLlxe4. Clearly, placing the bishop
on d3 makes more sense.
d) After 8 �d3 b5, 9 h3 transposes
to Line B 1 2, while 9 �f4 is note 'c' to
White's 9th move in Line B 2 1 3 of
Chapter 2. However, to limit Black's
options, White may as well play 8 h3
first.
8 1 1)
8 a4 �g4
This idea - ... �g4 and . . . �xf3, fol­
lowed by restraint on e5 - has been a
terrifically successful idea against
White's orthodox lines. If White plays
9 �e2, for example, we will reach a
position that could have come directly
from the Classical main lines of the
next chapter. Black feels that his knights
will be at least a match for White's
bishops, and with the c8-bishop out of
the way, it's easy to coordinate pieces
and connect rooks. Finally, Black can
still play for ... b5 (supported by ... l:tb8
and ...lLle8-c7, for example). For his
part, White has both more space and
the bishop-pair (an advantage, to be
sure), but finds himself searching for a
plan beyond general restraint.
Now there are three main moves:
Bll1: 9 h3 162
B112: 9 �e2 1 64
BIB: 9 'ili'b3 1 68
8 1 1 1)
9 h3 �xf3 10 �xf3 (D)
B
This somewhat misplaces White's
queen, making it vulnerable to attack
by . . . lLlbd7-e5. On the other hand, 1 0
�xf3 keeps the light-squared bishop
on its best diagonal and therefore de- .
serves close attention:
10 �g7
•••
The other option is 1 0. . . lLlbd7,
when 1 1 �f4 (after 1 1 �dl ! , Black
should transpose to the main line by
1 l . ..�g7, since 1 l . ..'i1ie7?! 1 2 �d3 c4
1 3 �xc4 lLlxe4 14 0-0 favours White)
1 l .. .�c7 1 2 �d3 �g7 1 3 0-0 0-0 in",
tending ...lLlh5 and ... lLle5 (or ...l:tfe8
and ... lLle5) ensures a level game.
l 1 �d3
This is the only real test. Other
moves:
a) 1 1 g4 ! ? 0-0 1 2 �g2 lLlbd7 1 3
'i1ig3 lLle8 1 4 0-0 c4 ! 1 5 g5 lLle5 =
 MODERN MAIN LINE
Douven-Antonio jr, Thessaloniki OL
1 98 8.
b) 1 1 i.e2 0-0 12 0-0 ttJbd7 is
harmless; e.g., 13 i.f4 ( 1 3 i.d2 1Ic8 !
14 1:.ael c4 I S g4 �b6 1 6 i.c 1 ttJeS
{ 1 6 .. :iHb4 ! ? } 1 7 'ilVe3 �xe3 1 8 i.xe3
lbfd7 ! = intending ... ttJcS, Kanko­
Inala, Finnish Cht 1 995/6) 1 3 ...'iVe7
14 1:.fe 1 ttJhS ! I S i.h2 i.d4 ! (that
theme again !) 1 6 g4 ! ? ttJhf6 1 7 i.n
hS ! 18 'iVg3 ttJeS 19 i.e2 hxg4 20 hxg4
gS !, securing f4 with advantage, Ap­
pel-Kapengut, Budapest 1 989.
11 ...0-0
After l l .. .ttJbd7, 1 2 'ilVdl amounts
to the same thing, while 1 2 i.f4 �e7
13 i.c2 0-0 1 4 0-0 allows 1 4 ...ttJhS !
with the plausible continuation I S
i.h2 ( 1 S i.d2 ttJeS 1 6 �e2 'iVh4 ! =)
IS . . . .i.d4 ! ? ( 1 S ... l:.fe8) 16 ttJe2 i.eS
(or 1 6 . . .ttJeS 1 7 'ilVb3 c4 1 8 'iVa2 i.a7
19 .i.xeS 'iYxeS 20 �xc4 1:.ac8 2 1
'ti'b3 fS ! with plenty for a pawn) 1 7
..txeS ttJxeS 1 8 'iWc3 bS =.
12 0-0
1 2 iLf4 ttJe8 (or 1 2. . :�lke7 1 3 0-0
ttJbd7, when 1 4 i.c2 transposes to the
note to Black's 1 1 th move, while Kap­
engut describes 14 �g3 ttJe8 as "un­
clear") 1 3 0-0 ttJd7 1 4 'iYe2 ! ? l:.b8- I S
a5 ttJeS 1 6 i.c2 ttJc7 1 7 i.d2 ( 1 7 .i.h2
b5 1 8 axb6 lhb6 19 l:.fb1 'iYb8 = )
17 ...1:.e8 1 8 i.a4 ttJd7 19 'iVg4 ttJe5 20
'ti'e2 ttJd7 2 1 'iVg4 l:.e7 ! ? 22 i.g5 f6 23
.tf4 ttJeS 24 'iVe2 b5 25 axb6 l:.xb6
with obscure prospects (Black might
play for ...'iVb8 and ...ttJbS-d4), Vojska­
Garcia Palermo, Saragossa 1 993.
12. ttJbd7 (D)
..
One could also try a different set-up
that waits for .i.f4 and emphasizes the
163
dark squares via 12 ...1:.e8 ! ? 1 3 iLf4
'iile7 1 4 l:.fe 1 ttJfd7 ! ; for example, 1 5
i.n ttJe5 1 6 �d 1 ( 1 6 'iiig3 ttJbd7 1 7 as
h6 =) 16 ...gS ! ? 17 i.e3 'iYf6 ( 1 7...ttJbd7
1 8 i.e2 h6 is equal) 1 8 f3 ttJbd7 1 9 as
ttJg6 intending . . . ttJdeS and ... ttJf4.
The text-move is more trustworthy.
w
13 ikdl!
Schneider calls this "artificial" and
gives it a ' ? ' , but I think it is White's
most logical move, since 13 'ilVe2 leads
to problems on the e-file, and 1 3 'iVg3
ttJhS ! poses Black no problems.
13 i:tc8
•..
Similar is 1 3 . . .l:te8 1 4 lIe l l:.c8 ! (I
discovered the dangers in this type of
position after 1 4 ...'iVc7? ! I S i.e3 ! c4
16 i.c2 ttJcS 17 i.d4 'iYaS 1 8 l:.e3 !
l:.ac8 1 9 l:.a3 h6? ! 20 f4 ± Sulypa­
J.Watson, Linares 1999) IS as ( 1 S .i.gS
c4 16 i.n h6 17 .i.h4 'iYb6 1 8 'ii'c2
ttJhS ! was already excellent for Black
in Novikov-Moiseev, Berlin 1 992)
IS ...c4 1 6 i.bl ( 1 6 i.c2 l:.cS ! ? 17 l:.a4
l:txaS 1 8 l:.xc4 bS 1 9 1:.b4 l:.a1 =)
1 6... ttJcS 17 i.c2 ttJfd7 1 8 f4 'ilVh4 !
with active play and the advantage,
164 THE GAMBIT GUIDE TO THE MODERN BENONI
Miles-de Firmian, USA Ch (Long
Beach) 1 989.
14 l:el
14 as c4 at best transposes.
14 c4 15 i.c2
•.•
IS .lin l:.e8 1 6 i.f4 ttJcS 17 'iVc2
ttJb3 1 8 l:.ad l 'Wic7, Granda-Hulak,
Zagreb IZ 1 987, is already fine for
Black in view of 1 9 eS? ! dxeS 20 d6
'iVcs 2 1 ttJe4 ttJd4 ! + (Granda).
After the text-move ( 1 S iLc2),
Mandekic-Hulak, Pula 1 993 contin­
ued I S . . .ttJcS 1 6 as ttJfd7 17 f4 (to
stop ... ttJeS-d3) 1 7 ... z:te8 1 8 l:ta3 'iVc7
(also promising is 1 8 ...ttJd3 ! ? 19 i.xd3
cxd3 20 'Wixd3 ttJcS 2 1 'iVd 1 iLxc3 ! 22
J:Ixc3 1:txe4, OJakobsen-Emms, Co­
penhagen 1 994) 1 9 eS? dxeS 20 d6
'iVc6 2 1 fS e4 +.
B1 12)
9 i.e2
A natural move. Now Black will ex­
change on f3 and then organize a re­
straining policy versus White's eS .
See the note to Black's 1 0th move re­
garding the transposition of this sys­
tem to the Classical Main Line.
9 i.xf3
•••
Black decides not to allow 1 0 ttJd2,
since he wants to make this exchange
anyway.
10 iLxf3 (D)
10 i.g7
.••
This is one of the key positions for
the 7 . . . a6 system, and also for the
Modern Benoni. After White's next
move ( 1 1 0-0), we have transposed di­
rectly to a variation that can arise via
the Classical Main Line of the next
chapter, i.e., 6 e4 g6 7 ttJf3 i.g7 8 i.e2
a6 9 a4 i.g4 10 0-0 i.xf3 1 1 iLxf3. It
is important to realize that the analysis
here applies to that chapter as well. In
fact, this is such an effective line that
one should happily employ it versus
the move-order of the Classical Main.
Line just mentioned. As we will see in
Chapter 1 0, White will often use other
move-orders involving an early ttJd2,
primarily to avoid having to face this
. . . iLg4 and . . . i.xf3 solution. Never­
theless, since the move-order given
above also arises regularly, that gives'
us all the more reason to learn it.
As a possible alternative for Black.,
1 O... ttJbd7 tends to lead to the same
position, except in the case of 1 1 i.f4
'iVb8 ! ? 1 2 0-0 iLg7 1 3 i.e2 0-0 14
'iVc2 J::!.e8 IS i.g3 ( I S h3 "WIc7 16 as
l:te7 17 l:!a4 J:tae8 1 8 I:.fal bS ! 1 9 axb6,
ttJxb6 20 l:.xa6 ttJxe4 = Rai.Garcia�1
Soos, Havana OL 1 966) I S . . .'Wic7 16
f4 c4 1 7 'it>h 1 1:tac8 1 8 as �d8 ! =, in�\
tending . . JkS and .. .l:haS, Wexler..
Bronstein, Mar del Plata 1 960.
11 0-0
1 1 "WIb3 b6 12 iLf4 0-0 1 3 0-0 ttJe8 :=
Sorin-de Firmian, Buenos Aires 1 995.
 MODERN MAIN LINE
11 0-0
...
Some may prefer playing l l ...lLlbd7
(to avoid 1 2 e5). This normally trans­
poses except in the case of 12 .i.f4
'fie7 1 3 e5 ! ? ( 1 3 :tel 0-0 transposes to
our main line) 1 3 ... dxe5 1 4 d6 �e6 1 5
lLld5 lLlxd5 1 6 .i.xd5 �f6 1 7 .i.g3 :b8
18 f4 e4! 19 .i.xe4 0-0, Calego-Franco,
Linares 1 99 1 , which Psakhis says is
"unclear". Black appears to have some
trouble coping with the bishops, but
20 l:tb l 'iVd4+ 2 1 'iVxd4 .i.xd4+ 22
t;t>h l ! ? (22 .i.f2 lUe8 =) 22 . . . lLlb6!
(covering d5) 23 f5 (23 b3 f5 24 .i.f3
l:tfd8) 23 ...lLlxa4 24 d7 l:i.bd8 25 fxg6
hxg6 26 .i.d6 ':xd7 is at least adequate.
12 ..1tf4
The main line in this much-played
position. Others:
a) 12 e5 ! ? dxe5 1 3 .i.g5 ( 1 3 d6 lLlc6)
13 ...lLlbd7 1 4 a5 ! h6 ! 1 5 .i.h4 �b8 1 6
d6 g5 1 7 .i.g3 lLle8 + intending .. .f5,
Y.Mikhalevski-Yudasin, Beersheba
1992.
b) 1 2 .i.g5 i:te8 ( 1 2 ... h6 1 3 .i.h4
ct:Jbd7 forces White's bishop to com­
mit sooner) 1 3 'iWc2 lLlbd7 14 .i.e2 h6
15 ..Ith4 'iVa5 ! 1 6 'it>h l !tac8 17 f3 c4 =
Beliavsky-Suba, Tunis IZ 1 985.
c) 12 a5 lLlbd7 13 :ta4 ! ? (or 1 3
.tf4 'iile7, when 14 l:.el transposes to
the main line, 1 4 'iVd2 to note 'b' to
White's 1 3th move, 14 �c2 to note
'c2' to White's 1 3th move and 1 4 �b3
to note 'dl ' to White's 1 3 th move)
13 . . . lLle8 14 .i.e2 lLlc7 15 'iVc2 'iWe7 1 6
.1d2 !tfb8 ! = 1 7 b 3 b 5 1 8 axb6 �xb6
with the idea of ...!tab8 and ... lLlb5-d4,
Pinter-Psakhis, Paris 1 990.
d) 12 'iWc2 lLlbd7 1 3 .i.d2 ! ? was
successful for White in Pinter-Emms,
165
Copenhagen 1 995 after 1 3 ...!te8 14 a5
:c8 1 5 l::tfe 1 c4 16 i..e2 lLle5 1 7 lLla4,
but I prefer the standard plan chosen
by Wojtkiewicz in a similar position:
1 3 ...�e7 14 a5 lLle8 ! 1 5 l::tfe l lLlc7 1 6
:a3 l::tab8 1 7 .i.e2 l:.fe8 1 8 .i. f l b 5 1 9
axb6 1:txb6 =.
12 :iVe7 (D)
..
w
13 �el
Considered best. Other moves are
interesting (or critical, in the case of
1 3 e5), but none of them threaten the
soundness of Black's system:
a) 1 3 e5 dxe5 14 d6 'iVe6 1 5 l1el
(after 15 .i.g5, 15 ...l:!.d8 ! = looks rather
easier than 1 5 ...lLlbd7 1 6 .i.xb7 :a7 =,
as has been played) 1 5 . . .lLlbd7 1 6
..Itxb7 l:ta7 ! ( l 6 . . . l:.ab8 ! ? 1 7 .i.xa6
1:!xb2 1 8 ..Itb5, Hjartarson-de Firmian,
Reykjavik 1 984, and now Kapengut
suggests 1 8 .. Jib4 ! ?) 1 7 .i.c6 ( 1 7 .i.f3
�b8 1 8 lLld5 'ii'xd6 1 9 .i.g5 lLlxd5 20
..Itxd5 'iVf8 ! is slightly better for Black,
Browne-D.Gurevich, New York 1 984)
17 ....l:!.b8 1 8 'iVd2 ( 1 8 ..Itxe5 ? lLlxe5 1 9
f4 lLlxc6! 20 l::txe6 fxe6 + Vermiash­
Pinchuk, COIT. 1997) 18 ...l'.:tb6 ! 19 lLld5,
166 THE GAMBIT GUIDE TO THE MODERN BENONI
C.Hansen-Y.Grtinfeld, Thessaloniki
OL 1 9S4, and here just 1 9, , :�xd6 ! ap­
pears very strong, as 20 ttJxb6 'iVxc6
leaves White with two pieces hanging.
b) 1 3 'iVd2 can be met in several
ways, but an instructive example arose
from 1 3 ".ttJbd7 1 4 a5 h5 ! ? (a recur­
ring theme in the Benoni, especially
when White has bishops on f3 and f4 -
Black will often play " .ttJh7, in order
to threaten ".g5-g4, but he also has
".ttJg5 in some cases; a more conven­
tional equalizer is 1 4" J:tabS 1 5 ..ie2
ttJeS intending ".ttJc7 and " .b5) 1 5 h3
(to prevent ".ttJg4, another idea be­
hind ".h5) 1 5 . . . J::!. abS ( l 5 " .lhcS 1 6
J::tfe 1 c4 1 7 ttJa4 ttJe5 I S .ixe5 'iVxe5
was equal in Gaprindashvili-Madl,
Smederevska Palanka worn IZ 1 9S7)
1 6 J::!.a 2 ttJh7 1 7 J::!.e 1 J::!.feS I S i.e2
i.d4 ! 1 9 Ma4 ttJhf6 = Ruban-Emms,
Hastings 1 99 1 12.
c) 1 3 'iVc2 ttJbd7 and now:
c l ) 1 4 i.e2 ttJeS 1 5 a5 ttJc7 1 6
J::!.fe l J::!.abS ( a familiar manoeuvre ! ) 1 7
J::!.a4 b5 I S axb6 J::!.xb6 = Velikov-Er­
menkov, Albena 1 9S3.
c2) 14 a5 ttJe5 1 5 i.e2 ttJfd7 1 6
.ie3, Lukacs-Groszpeter, Kecskemet
1 9S3, and now Black might try the un­
clear 1 6".g5 ! ?
c3) 1 4 h3 l:i.abS 1 5 a5 ttJeS 1 6 J::!.fe l
ttJc7 1 7 i.e2 b5 = Krush-Scherbakov,
Presov 2000.
c4) 1 4 J::!.ae l c4 ! ? 1 5 i.e2 l::tacS 1 6
.Jtg5 h6 1 7 .Jth4 g5 ! ? ( 1 7 ".l:i.feS, in­
tending I S f4 b5 ! ?) I S i.g3 l:lJeS =
Agzamov-Psakhis, Baku 1 979; on 1 9
a5, simplest i s 1 9 " .ttJe5 .
d) 1 3 'iVb3 ! ? has been tested re­
cently: 1 3 " .ttJbd7 ! and then:
d l ) 1 4 a5 J::!.abS 1 5 ttJa4 h5 ! 1 6
ttJb6 ttJeS with comfortable play for
Black, Roa Alonso-Topalov, Villarro­
bledo rpd 2000.
d2) 1 4 'iVxb7 J::!.fbS 1 5 'iVc6 ttJeS ! ?
( l 5 ".ttJe5 1 6 .ixe5 'iVxe5 "with some
compensation" - Emms) 1 6 i.g4 ( 1 6
a5 i.e5 ! Emms) 1 6" .ttJe5 1 7 .Jtxe5
..ixe5 I S f4 ..id4+ 1 9 �h I 'iVdS 20 e5
f5 ! 2 1 exf6 (21 .if3 J::!.b6 -+) 2 1 .. .ttJxf6
22 .ie6+ WhS 23 a5 J::!. xb2 24 ttJa4 and
now 24 . . . J::!.b bS, as played in Pilar­
Meduna, Czech League 1 997/S, is OK,
but 24".Mb5 ! is more active and looks
somewhat better for Black.
13 ttJbd7 (D)
...
14 as
Or:
a) 1 4 h3 is well-suited to counter
" .h5, but it is also slow and can be an­
swered by the standard plan to prepare
".b5: 1 4" .J::tabS 1 5 ..ie2 ttJeS 1 6 'i¥d2
ttJc7 1 7 .ig5 ! ? .if6 I S .ih6 J::!.feS =
Portisch-Kindermann, Dubai OL 1 9S6.
b) 1 4 'iVd2 lUeS (again, the dy­
namic 14".h5 100ks quite playable: 1 5
..Ith6 ttJe5 1 6 i.e2 .ixh6 1 7 'iVxh6
 MODERN MAIN LINE
lbfg4 18 i.xg4 hxg4 19 h3 Wif6 =
Sahovic-Vera, Aosta 1 989) and now:
b l ) I S h3 c4 ( 1 S ...tDeS I 6 ..ite2 'iVc7
1 7 i.h2 c4 1 8 'iith 1 .l:!.ab8 1 9 f4 tDed7
20 i.f3 bS 2 1 axbS axbS = Ehlvest-de
Firmian, New York 1 997) 16 .ie2
nac8 17 .if! tDeS 1 8 ..ith6 .ixh6 19
'i'xh6 "fiIc7 20 'iVd2 tDed7 2 1 as Wid8 =
intending . . . l:tcSxaS, Gligoric-Suba,
Vrsac 1983.
b2) IS as hS ! 1 6 h3 .l::!.ac8 (or again,
1 6 ... tDh7 ; e.g., 1 7 lIa3 �f6 1 8 tDa4 gS
19 iLh2 "fiIg6 20 i.dl .id4 21 i.c2
ttJeS intending . . . g4 or just . . .tDf6,
Hjartarson-Suba, Manila OL 1 992)
and then:
b2 1 ) 17 b3 tDh7 ! 1 8 1:.ac l ..itd4 1 9
i.e3 �f6! 20 ..itf4 tDgS (20. . . ..itgS ! =
looks more accurate, leaving White
with some structural problems and a
bad bishop) and now, rather than 2 1
'i'e2? tDeS 22 ..itxeS .ixeS + 23 tDdl ?
ttJxh3+ -+ Henrichsen-Emrns, Copen­
hagen 1 993, Graham Burgess notes
that 2 1 i.d l ! ultimately works out tac­
tically if Black grabs the e-pawn by
2 l . . .�xc3? ! 22 .l:Ixc3 tDxe4? 23 l:tce3.
Still, 2 1 ..:iM8 ! isn't bad, when the e­
pawn really is threatened, and 22 f3
'i'e7 = gives Black dark-square weak­
nesses to work on after . . .tDh7.
b22) After 17 .idl !, 17 ...c4 1 8 ..itc2
ttJeS 1 9 i.e3 tDfd7 20 tDa4 tDd3 ! 2 1
i.xd3 cxd3 22 tDb6 tDxb6 23 .ixb6
:tc2 24 Wixd3 1:.ec8 yielded compen­
sation in C.Hansen-Ward, Copenha­
gen 1 996. 17 . . . tDh7 should also be
considered, as in line 'b2 1 ' .
14 h5 (D)
...
Just about every possibility has
been tried here, mostly without many
167
problems for Black. Two instructive
lines:
a) 14 .. JHe8 and then:
a l ) I S h3 ! ? can be answered by
I S .. J:!.ac8!, intending . . . c4. Instead,
I S . . .1:.ab8 1 6 ..ite2 bS 1 7 axb6 .l:i.xb6 1 8
Itxa6 tDxe4 1 9 ..itd3 ..itxc3 20 bxc3
.l:i.xa6 2 1 i.xa6 'iVh4 22 'iVf3 fS was
unclear and probably equal in Ger­
shon-Marin, Tel-Aviv 2000.
a2) I S 'iVd2 hS ! ? 1 6 h3 tDh7 1 7
l:ta3 �f6 1 8 tDa4 gS 1 9 i.h2 Wig6 =
Hj artarson-Suba, Manila OL 1 992,
based upon 20 Widl tDhf6 ! 21 .ixd6
g4 22 hxg4 hxg4 23 ..ite2 tDxe4 =.
b) 1 4 ... tDeS I S .ie2 l:lfe8 1 6 ..itg3
gS ! 1 7 .l:!.a4 ( 1 7 'iVd2 ! ? intending tDdl ­
e3-fS) 1 7 . . . tDfd7 1 8 .if! tDg6 with
dark-square control, Barlov-de Firmian,
Novi Sad OL 1 990.
15 tDa4
Or:
a) I S g3!? nab8 1 6 tDa4 tDe8 1 7
l:tc 1 'iVd8 ! 1 8 'il¥d2 bS 1 9 axb6 tDxb6 =
Gligoric-Psakhis, Sarajevo 1 986. I
prefer Black's dynamism after the
game continuation 20 tDxcS ! ? dxcS 2 1
1 68 THE GAMBIT GUIDE TO THE MODERN BENONI
�xb8 'iYxb8 22 J:IxcS CLld6, although
this may be a matter of taste.
b) I S h3 can be answered by the
ever-reliable I S . . . l:i.ab8 1 6 ..Ite2 CLle8
1 7 'iYd2 CLlc7 ; e.g., 1 8 ..Itd3 bS 19 axb6
Mxb6 20 l:i.a2 Mtb8 =, after which 2 1
Meal Mb4 22 'iYc 1 CLleS 23 ..Ite2 CLlc4 =
might follow.
IS CLlh7
..•
Black threatens . . . gS-g4.
16 ..Itd2 CLlgS !
Showing yet another idea behind
the . . . hS, . . . CLlh7 manoeuvre: one of
White's bishops is traded, or the e­
pawn falls.
Now:
a) 1 7 g3 CLlxf3+ 112-112 Beliavsky­
Psakhis, USSR Ch (Minsk) 1 987.
Black's position seems preferable.
b) 1 7 �e2 CLlxe4 1 8 ..Itd3 fS 1 9 f3
�d4+ ( l 9 ... CLleS ! ? 20 �xe4 fxe4 2 1
Mxe4 'iYf7 22 CLlb6 l:i.ae8 also looks
OK) 20 i.e3 �xe3+ 2 1 Mxe3 'iYg7 ! 22
�xe4 (22 fxe4 'iYd4 -+) 22 . . . l:i.ae8 !
with full equality.
This has been a long overview of a
crucial variation that can be used
against both the Modern and Classical
main lines of the Benoni. By review­
ing this section, one can see why White
has for the most part been avoiding the
. . . �g4 and . . . �xf3 variation. Black's
blockade on eS is firm, and he has
chances on both sides of the board. In
particular, the idea of . . . hS and . . .CLlh7
has confirmed that White has practical
difficulties in a variation which only
promises equality anyway.
8 1 13)
9 'iYb3 (D)
This is White's most tactical and
forcing move, which virtually wins a
pawn by force, since 9 . . . �c8 is pa­
thetic and 9 . . .'iYc7 10 CLld2 ! prepares
f3, CLlc4 and i.f4. Then Black has lost
two tempi over normal lines. This
leaves:
9 �xf3 10 'iYxb7
...
1 0 gxf3 'iYe7 ! ? (a straightforward
solution is just 1 O . . . 'iYc7 ! 1 1 as CLlbd7
with easy play) 1 1 �f4 CLlhS 1 2 �xd6
'iYxd6 1 3 'iYxb7 ..Ith6 ! 14 �bS+! 'it'f8
I S i.e2 ct;g7 1 6 'iYxa8 �b6 1 7 J:lbl
CLlf4 with dark-square control and a lot
of (but not necessarily enough) com­
pensation, Molvig-Emms, Copenha­
gen 1 996.
10 �xg2
...
This position is a joy to analyse,
and a paradigm of attacking chess.
The text-move is the straightforward
approach, but 1 O . . . CLlbd7 can be great
fun and is probably the best move. 1 1
gxf3 �g7 (D) and then:
a) 1 2 �xa6 J:lb8 1 3 'iYc6 0-0 1 4 as
( 1 4 'iYxd6? l:i.b6) 1 4 . . . CLleS I S �a4
CLlxf3+ 1 6 ct;fl CLlg4 ! 1 7 h3 ! ? 'iYh4 1 8
�c2 CLld4 with a huge attack.
 MODERN MAIN LINE
b) 1 2 'iVb3 lbhS ! (with the idea
...�h4) 1 3 h4 i.eS 14 i.gS �aS I S
i.d2 i.f4 1 6 'ilic2 �d8 1 7 lbd l lbeS,
again with complete dark-square con­
trol, Kurz-Emms, Baden-Baden 1 992.
c) 12 i.f4 ! ? (this position can arise
in several ways) 1 2 . . .J::tb 8 1 3 'iYxa6
J:txb2 ! (D) (I find this move fascinat­
ing, and it may even be objectively
better than 1 3 . . .lbhS 14 i.c 1 ! 0-0 I S
i.e2 i.eS with compensation, Van
Wely-Topalov, Amsterdam 1 997).
w d) 1 2 �c6! 0-0 1 3 'iixd6 lbhS and
Here are some engaging possibili­
ties:
169
c 1 ) 14 i.bS lbhS IS i.d2 ( 1 S
i.xd7+ 'tWxd7 1 6 i.d2 0-0 intending
moves like . . . 'tWh3, . . . i.d4 and . . . fS)
l S . . .0-0 16 'tIi'xd6?? (but 1 6 Ita3 lbeS
1 7 lbd1 �xd2 ! 1 8 �xd2 'iVgS+ 1 9
Wc2 'tIi'g2 20 lIe1 lbxf3 2 1 J::txf3 �xf3
22 'ilfxd6 i.d4! with the idea . . .'ilia3 fa­
vours Black) 1 6 . . .�aS ! 1 7 lbb1 ( 1 7
"fia6 'ilfb4 1 8 �c 1 lbeS ! -+) 1 7 . . .�xd2
1 8 lbxd2 i.xa1 1 9 �e2 .i.eS 20 'ilixd7
lbf4+ -+.
c2) 1 4 lbbS 0-0 I S i.e2 ( 1 S �d1 ?
lbxe4 ! 1 6 fxe4 �h4 1 7 i.e3 'ifxe4 1 8
J:tg1 'ilib4+ with a killing attack; I S
.i.gS �e8 1 6 i.g2 h6 1 7 i.xf6 { 1 7
i.e3? lbxdS ! -+ } 1 7 ....i.xf6 1 8 0-0
i.eS +, for example, 1 9 'iVc6? lbb8 ! 20
"fic7 'ilih4 2 1 h3 'tWf4 22 l:.fb1 �c2 !
-+) l s . . .lbxdS (or l s . . . lbhS; e.g., 1 6
.i.xd6?! �f6 { threatening . . .lbf4 } 1 7
'tWaS �xbS 1 8 axbS 'iVxd6 with terrific
activity for Black's minor pieces) 1 6
exdS "fih4 1 7 'tWxd6 l:!.e8 1 8 0-0 ( 1 8
.i.d2 �xd2 19 �xd2 �+) 1 8 ...�bxe2
1 9 lbc7 ! i.eS 20 .i.xeS l:.8xeS ! 2 1
'tWxd7 J:.gS+ 22 �h 1 IlhS 23 'iWc8+
�g7 24 h3 1.lfS 2S �g2 J:.gS+ 26 �h 1
1.lxf2 -+.
now:
d 1 ) 14 i.h3 is given as ' ! ±' by
Psakhis and Kapengut, but Schneider
refutes that assessment by 14 ...fS I S
1i'e6+ �h8 1 6 d6 lbeS ! 1 7 �e2 �h4
1 8 i.g2 lbd3 ! (Schneider also claims
an "immediate win" by 1 8 . . .fxe4 1 9
lbxe4 :ae8 20 �dS lbg4 2 1 'iixcs
�xe4+, but 22 fxe4 ! is at best unclear)
1 9 �xd3 'iWxf2 "+ with a powerful at­
tack". Then 20 �d l �xg2 is complex,
but promising for Black.
1 70 THE GAMBIT GUIDE TO THE MODERN BENONI
d2) Gavrikov-Lutikov, Minsk 198 1
went 14 f4 Ite8 IS .ig2 lta7 ! (threat-
ening 1 6 ....if8 1 7 1IYc6 liJb8, and pre-
paring . . .'ii'M ) 1 6 as 'ii'M 1 7 liJa4
l:!.c8 ! 1 8 liJb6 .if8 '+' Kapengut, and
indeed, 19 liJxc8 .ixd6 20 liJxd6 liJxf4
is quite good for Black.
11 .ixg2 liJbd7 (D)
w 14 ...dxeS IS d6 .ixd6 1 6 'ii'xa8 'ii'xa8
12 0-0
Probably best. Otherwise:
a) 1 2 'i¥c6 was long considered
good, but then 1 2 ....ie7 ! is at least ad­
equate; e.g., 1 3 eS ! ? ( 1 3 '.if4 �b8 14
.ih3 l'h7 =) 13 ....l:f.c8 1 4 exf6? ( 1 4
'ii'xa6 liJxeS = ) 1 4 ... ltxc6 I S fxe7
'iVxe7+ 1 6 �d l .l:f.b6 and Black wins,
Olafsson-de Firrnian, Reykjavik 1 994.
b) 1 2 .if4?! liJhS 1 3 .ih3 fS 1 4
.igS ltb8 ! I S .ixd8 l:Xxb7 1 6 .iaS liJeS
1 7 .i n , Najdorf-Camarra, Mar del
Plata 1 96 1 . This must be the stem game
for 9 'iVb3. Now Kapengut suggests
17 ...l:lxb2 1 8 l:.a2 l:!.xa2 1 9 liJxa2, and
here instead of his 1 9 ... .ih6?, Schnei­
der proposes 1 9 ... fxe4 ! 20 .ixa6 liJf6
(20. . .�e7 ! ?) 2 1 liJc3 .ih6 22 .ibS+
�e7 +.
c) 12 .ih3 liJeS 13 0-0 (Vezzosi­
Gardon, Toscana 1 999) 1 3 .. :iVb8 ! '+.
12 1IYb8!?
.••
Perhaps not thrilled with this move
(see the note to White's 1 6th move, for
example), Black has often turned to
1 2 ... .ie7 of late: 1 3 f4 0-0 1 4 eS (the
recent 14 as lte8 { 14...c4 ! ? } I S 1IYc6
can be answered by I S .. :iVb8 ! ; e.g., 1 6
l:!.a2 11Ya7 1 7 �h I lbc8 1 8 'iVa4 .if8 =,
having in mind both ... .ig7 and ...c4)
17 .ixa8 l:.xa8 1 8 as ( 1 8 fxeS .ixeS
19 .if4 .id4+ was equal in Murdzia­
Yakovich, Swidnica 1 999) 1 8 . . . lte8
19 :a4 �g7 20 Itd l :e6 '+ (White's
pawns are loose) Crouch-Emms, Brit­
ish Ch (Millfield) 2000. 1 2 ... .ie7 of­
fers greater winning chances than
1 2 ...'ii'b 8, as we shall see.
13 �c6 .ig7 14 .if4 �e7!
Threatening .. �:c8.
15 e5 dxe5 (D)
w
16 :fel!
White can win the exchange but
achieve little by 1 6 d6+ �xd6 1 7
�xd6+ �xd6 1 8 :fd l + �c7 1 9 .ixa8
 MODERN MAIN LINE
lha8 =. More interesting is the un­
tested 16 .tgS, which comes close to
gaining the advantage, although it's
not clear how after 1 6 ... l:ta7 1 7 d6+
�xd6 1 8 ltJdS+ <it>e6 1 9 .th3+ ltJg4
20 ltJe7 ! (20 .txg4+ fS) 20 . . .fS 2 1
I:tfd 1 �xc6 22 ltJxc6 �c7 23 ltJd8+
Mxd8 24 .txd8 ':'c8, when it appears
that Black has enough compensation
for the exchange. This kind of thing
may well be the reason players are
turning to 1 2 . . . .te7, however.
16.. J�dS! 17 d6+
Komarov analyses 1 7 .txeS ltJxeS
18 f4 ltJfd7 1 9 d6+! �xd6 ! 20 ltJdS+
We6 2 1 ltJc7+ <it>e7 = with the point
that 22 �xd6+? loses to 22 . . .<it>xd6 23
ltJxa8 ltJc6! -+.
After the text-move ( 1 7 d6+), Kom­
arov -Foisor, Lyons 1 995 continued
17 . . . 1Wxd6 1 8 �xd6+ <it>xd6 1 9 l:1.ad 1 +
We6 2 0 .tgS ! ? h 6 2 1 ltJdS (the alter­
native 2 1 .txf6 ltJxf6 22 .txa8 1J.xa8
gives Black two pawns for the ex­
change and counterplay via . . J:I.b8)
2 l .. .hxgS 22 ltJc7+ <it>e7 23 .txa8 (23
ltJxa8 :b8 !) 23 .. ..lIc8! 24 ltJdS+ ltJxdS
2S i.xdS ltJb6 ! (or 2S ...fS = with an
aggressive centre) 26 as ltJa4; this is
still unclear - a great struggle.
But for more positive chances, Black
should deviate from the main line
given here by 1O ...ltJbd7 (which I think
is objectively superior to 10 ... .txg2),
or 1 2 . . . .te7. Generally speaking, I
would be surprised if White continues
to enter into these 9 'i!Vb3 lines unless
he can improve at an early stage.
8 1 2)
S h3
1 71
Even if 8 i.d3 bS 9 h3 leads to the
same position, there's no reason to al­
low Black to deviate on move 8.
S b5 9 .td3
•..
Black threatened . . .b4 and . . .ltJxe4.
9 ... .tg7 10 0-0 (D)
B
10...0-0
A key position, one which theory
has held to be perfectly good for Black
- see, e.g., NCO, MCO, ECO, Kapen­
gut, Psakhis, and Schneider (who even
gives White a ' ? ! ' for entering the po­
sition). But recently, White has been
making progress in this line, espe­
cially against the known remedies, so
the variation deserves a detailed ex­
amination.
In my opinion, Black should not in
general play the move . . .c4 early on, at
the least not before he has all his mi­
nor pieces and a rook in play. The cost
of giving up the d4-square seems too
great in most cases. I'm sure that there
are exceptions to this, but I haven't
found a convincing one.
On the positive side, Black seems to
do quite well by leaving the pawns on
1 72 THE GAMBIT GUIDE TO THE MODERN BENONI
bS and cS for a while as he develops
his pieces. In principle, a drawback
might be the move b4, which attempts
to block Black' s queenside advance.
But in practice, White is seldom suc­
cessful with that move, if only be­
cause it helps Black to open lines on
the side of the board where he is stron­
ger. Thus, b4 may be considered about
an even bargain at best, and one which
costs White time to implement.
I should also point out that Black
makes the rather ugly move . . ..i.b7 in
most games involving this variation.
It's true that the bishop normally has
nowhere else to go, but it also serves
an important prophylactic function
versus eS, White's most natural break.
The proper disposition of Black's rooks
tends to be a more complicated ques­
tion: they may end up on c8 or b8 (sup­
porting a queenside pawn advance),
on d8 (guarding against eS), or on e8,
exerting pressure down the e-file.
In line with the explanation above,
1O ...c4 I I .i.c2 0-0 is unnecessarily
committal, ceding d4 to a bishop or
knight before developing. Black can
always play this later.
l l .i.f4
A difficult choice, since the same
position can arise by several move­
orders, and yet Black can react differ­
ently to each:
a) I I l:f.el is somewhat less direct
than I I .i.f4, but it is still something
that White will play soon in most
lines: 1 1 . . .ttJbd7 (a move that is pre­
vented by .i.f4; I l .. J:te8 1 2 .i.f4 trans­
poses to the main line), and now White
has a choice:
a l ) 1 2 .ltf4 ttJe8 (an option not
available to Black in some other lines;
1 2 . . .�b6 transposes to the main line)
13 "iid2 ( 1 3 .i.gS ! ? .ltf6 14 .i.h6 .ltg7
I S .ltxg7 �xg7 16 a3 ttJeS 17 .ltfl is
slightly better for White, Sammalvuo­
Petersen, Copenhagen 1 998) 1 3 . . . .i.b7
1 4 .i.h6 ( 1 4 a3 ! ?) 1 4 ... .ltxh6 I S �xh6
�f6 1 6 .ltfl ! ? "iig7 1 7 "iid2 l:Id8 1 8 a3
ttJef6 = O.1akobsen-Moskalenko, Co­
penhagen 1 995. Prophylaxis versus
eS .
a2) 1 2 a4 ! ? b4 ( l 2 ... c4?! 1 3 .i.fl
b4 14 ttJa2) 13 ttJbl as ! 14 ttJbd2 ( 1 4
.ltf4 ! ? looks more promising at first,
because then ttJbd2 won't block the
bishop, but 14 ...ttJb6 threatening ... ttJh5
is fine) 14...ttJb6 I S �a2 .i.a6 1 6 .i.xa6
!:txa6 1 7 b3 :a7 1 8 .i.b2 �e8 1 9 �a l !?,
Psakhis-Spraggett, Barcelona 1 993,
and now Kapengut analyses 19 . . .ttJh5
20 .ltxg7 ttJxg7 2 1 ttJc4 ttJxc4, which
must be at least equal, with . . .lhe7 to
come.
b) 1 1 a3 (D).
B
A frequently-played move, both di­
rected against . . .b4 and preparing b4
 MODERN MAIN LINE
in some cases. Nevertheless, it still al­
lows .. .t2Jbd7 and a quick consolida­
tion on Black' s part. See these typical
and instructive examples:
b 1 ) 1 l . . .l1a7 ! ? 1 2 l:te 1 z:!.e7 may
transpose, via 1 3 i.f4 l1feS 14 WHd2, to
note 'b2' to Black's 1 2th move.
b2) l l .. .�eS 1 2 �e1 ltJbd7 1 3 i.f4
"iVb6 ( 1 3 ...c4 14 .i.c2 ltJcS IS �d2 i.b7
1 6 l::tad 1 �b6 1 7 i.h6 t Browne-de
Firmian, USA Ch (Key West) 1 994;
this seems to me the normal result of
playing an early ...c4) 14 �d2. Now
we have arrived at a representative po­
sition:
b2 1 ) The sort of thing I would like
to avoid is Yermolinsky-Wedberg, New
York 2000, which went 1 4. . .c4(?!) I S
.1i.c2 i.b7 1 6 i.e3 �c7 ( 1 6. . .�dS 1 7
i.d4 .l:!.cS l S ltJh2 hS ! ? t) 1 7 i.d4 (or
17 ltJh2 ! ? ltJcS ? ! I S i.d4 .i:tabS 1 9
Ct:Jg4 ! with advantage, Dreev-Emms,
Hastings 2000) 17 .. J:tacS I S Ct:Jh2 ltJcS
(Emms mentions I S ... i.aS, followed
by ...'iWbS and ... as, but I think some­
thing has already gone awry) 1 9 l:tad 1
i.aS 20 f4 'iWbS 2 1 ltJg4! ltJxg4 22
i.xg7 rJi;xg7 23 hxg4 ltJd7 24 �d4+
\tgS 2S gS .l:.e7 26 �d2 1:!.ceS 27 ltJd1
(contemplating ltJe3-g4) 27 ...�b6 2S
'iWxb6 ltJxb6 29 ltJc3 ltJd7 30 eS ! ltJcs
(30 . . . dxeS 3 1 d6 l:i.e6 32 fS !) 3 1 ltJe4
etJxe4 32 i.xe4 dxeS 33 d6 l:Id7 34
i.xaS ':'xaS 3S fxeS with a winning
position.
b22) Emms mentions the logical
and restrained 14 . . .i.b7 IS I;Iac 1 .:tacS.
Then 16 b4 transposes to the main
line.
b3) 1 1 . ..ltJbd7 (most accurate, I
think) 1 2 i.f4 and now:
1 73
b3 1 ) 12 ...ltJeS 13 �d2 i.b7 14 �fe1
f6 ! ? has been played, but isn't much
fun.
b32) Another idea without ... c4 is
12 . . .'iWe7 ! ? 1 3 l:i.e1 ltJhS 14 i.gS ( 1 4
i.h2 ltJeS) 14 ... i.f6 I S �e3 l:tbS 1 6
�d2 ltJeS 1 7 Ct:JxeS i.xeS I S i.fl b4 ! ?
( l S ... i.d7 !) 1 9 axb4 .l:txb4, Lazarev­
Poluliakhov, USSR 1990. In general, I
like this treatment, although I'm a lit­
tle sceptical of l S ...b4 because White
can now play 20 i.xa6 i.xa6 2 1 l:!.xa6
l:tfbS 22 :ta2 i.xc3 23 bxc3 I;Ixe4 24
l:i.ea1 with dark squares and activity.
b33) 1 2 ... �b6 and here:
b33 1 ) 1 3 b3, to stop . . .c4, can be
met by 1 3 ... ltJhS 14 i.d2 ltJeS.
b332) 13 :Le I and now:
b332 1 ) Another example of the
... c4 thrust is 13 ... c4?! 14 i.c2 ltJcs, as
played in Rai.Garcia-Rosito, Argen­
tine Ch (Saenz Pena City) 1 996, when
I S eS dxeS 1 6 i.xeS i.b7 1 7 �d4
�adS I S b4 (or I S l:.ad 1 ) l S . . . cxb3 1 9
i.xb3 looks good for White.
b3322) 1 3 . . . i.b7 can be met by a
plan to enforce b4, though after 14
�d2 -UacS I S i.e3 �dS ! 1 6 .l:tac 1 .l:l.eS
17 b4 cxb4 I S axb4 ltJb6 ! =, White's
c4 is as vulnerable as any weaknesses
Black himself has.
b3323) A simple reply is 13 ... ltJhS !?
14 i.h2 ltJeS I S ltJxeS �xeS 16 i.xeS
dxeS intending . . .ltJf4 and . . .fS at some
point.
b333) 1 3 �d2 i.b7 1 4 l:!.fe1 �acS;
e.g., I S eS? ! dxeS 16 ltJxeS ltJxeS 17
i.xeS .l:tfdS with a fine game. Note
that .. J:teS was never played, saving a
tempo in this particular case.
11 �e8
...
1 74 THE GAMBIT GUIDE TO THE MODERN BENONI
Without the possibility of . . .lbbd7,
Black plays for restraint. 1 1 . .JWb6 ap­
pears awfully early, but maybe 1 1 ...iLb7
1 2 .:te l �b6 is playable, trying to get
back to note 'b3322' to White's 1 1 th
move. This transposition would occur
after 1 3 a3 lbbd7, and I'm not con­
vinced that, say, 1 3 'iYd2 is any better;
then 1 3 . . . .l:le8 transposes to the note to
White's 1 3th move.
For 1 1 . ..l:.a7 ! ?, see the next note.
12 1:tel (D)
B initely be played) 16 iLe2 lbh5, Zhu
12 i.h7
...
1 2 . . ..l:h7 ! ? is a slightly eccentric but
appealing move played with success
by Kamsky and Scherbakov. It has
the idea of .. .lIae7 before developing
Black's other pieces. Most sources
treat 12 .. J:ta7 briefly or not at all:
a) Kapengut quotes Stohl's 'refu­
tation' 1 3 i.xbS ' ! ' axbS 1 4 lbxbS,
when the line continues 14 . . .�d7 I S
lbxd6 l:txd6 1 6 i.xd6 'iYxd6 1 7 eS ±.
But I think Black should play 14 ... lbhS
instead: I S lbxa7 lbxf4 1 6 lbxc8
\\Vxc8, with a tough position to assess.
To me, White's queenside pawns look
harmless, as in the Benko Gambit.
White should probably exploit the
awkwardness of the knight on f4 by 17
'iYd2, but after 1 7 . . . lbhS followed by
moves like ... lbd7, ... �a6, and . . . lbe5
or . . .lbhf6, I don' t see any advantage
at all for White. Also, Black may be
able to skirt the whole issue a move
earlier by playing 1 1 . . . .i:.a7.
Having said all that, we should look
at some actual examples of play in this
potentially important line:
b) 1 3 'iid2 .uae7 and now:
b l ) 1 4 iLgS and after 14 . . .lbbd7? !
I S 'i!Vf4 lbeS?! ( 1 S . . :i!Vc7 = should def­
Chen-Oll, Beijing 1 997, 1 7 iLxe7 ! has
to be advantageous. A simple solution
is 1 4 . . .b4 l s lbd1 �b6 =.
b2) 1 4 a3 �b6 I S iLgS ( 1 S :e2
lbbd7 1 6 �ae 1 lbhS 1 7 iLe3 iLd4! 1 8
i.c2 i.xe3 1 9 ':xe3 lbeS 20 lbxe5
.uxe5 liz-liz Piket-Kamsky, Amster­
dam 1 996) l s . . .lbbd7 1 6 �f4 lbhS I 7
'iid2 (Scherbakov gives the amazing
line 1 7 �h4 ? :eS ! ! 1 8 lbxeS lbxeS 1 9
iLe2 h6 20 iLe7 { 20 iLe3?? iLf6 }
20 .. :iWc7 2 1 iLxh5 gxhS 22 iLf6 lbg6
+) 1 7 . . .lbhf6 1 8 �f4 lbhS 1 9 'iWd2
liz-liz Mitenkov-Scherbakov, Moscow
1 999.
13 a3
Initiating a plan to block the queen­
side by h4. The main alternative, again,
is 1 3 'iid2 �b6 ( 1 3 . . .c4 14 iLc2 1\Vb6
1 5 i.h6 transposes to line 'c') and
now:
a) The trick 14 i.xd6 'iYxd6 IS eS
'iYd8 16 exf6 iLxf6 achieves little here:
1 7 l::txe8+ 'iYxe8 1 8 lbe4 lbd7 1 9 :el
'i!Vf8 and White had nothing in the
 MODERN MAIN LINE
game Kononenko-Gordon, Rimavska
Sobota 1 996.
b) After 14 a3, 14 . . .c4?! 15 i..c2
tDbd7 16 i..e3 transposes to note 'b2 1 '
to White's 1 1 th move, while 14 ... tDbd7
is line 'b22' of that note.
c) 1 4 i..h6 c4 (as usual, I'm not
happy with this move, although in this
case, Black had a better chance to hold
the balance; in my opinion, 14 . . .ctJbd7
is best, when Black's position is com­
pact and resilient; again, ...c4 can al­
ways be played later) 1 5 i..c2 i..h 8,
and we have arrived, by a very obscure
and probably non-optimal transposi­
tion (Black actually played . . .c4 on
move 1 O !), at Bates-Emms, British
League (4NCL) 1 999/00, which took
an original course: 1 6 ctJe2 ! ? ctJbd7
( 1 6 . . . ctJxe4 1 7 i..xe4 lhe4 1 8 ctJg3
lIxe 1 + 19 :i.xe 1 ctJd7 20 ctJf5 i..f6 2 1
CUe7 + i s too scary to contemplate) 1 7
tDg3 .l:tac8 (Yermolinsky proposes
17 ...ctJc5 as an improvement, continu­
ing 18 i..e3 ctJfd7 1 9 i..d4 i.. xd4 20
'iVxd4 ctJe5 =, but White should play
1 8 �ad l ! ;!;) 1 8 i..e3 ctJc5 1 9 i..d4 a5?
20 ctJf5 ! 1\Vd8 21 ctJh6+ �f8 22 ctJg5
and White was winning.
We now return to 1 3 a3 (D):
13 J�Vb6
•.
13 ... c4 14 i..c2 'iVb6 1 5 i..e3 1\Vc7
has been played as well; I like to keep
d4 covered for a while.
14 �cl ctJbd7
The familiar thematic position, but
with . . . l:te8 thrown in.
15 b4
Again this idea, which normally
gives Black as much play as White.
15 �ac8 16 'iVd2
•••
1 75
1 6 i..e3 1\Vd8 ! 1 7 'iWd2 cxb4 1 8 axb4
ctJb6! 1 9 i..g5 ( 1 9 i..xb6 'iUxb6 = )
1 9...ctJc4 20 1\Vf4 h6 ! with the point 2 1
i..xh6? ctJh5 22 1\Vg5 i..f6 2 3 'iVg4 ctJb2
-+; 1 6 bxc5 ctJxc5 1 7 i..b 1 ctJh5 1 8
i..d2 'iWa5 +.
16 ctJh5!? 17 i..h6 i..h8 18 g4!?
•••
Trading weaknesses for space.
18 ctJhf6 19 i..e3 h5!?
•••
O r 1 9 . . .1\Vd8 = , again with the idea
of . . . cxb4 and ... ctJb6-c4.
20 g5 ctJh7 21 ctJh4
This curious move prevents . . . f6.
Instead, 2 1 a4 is met by 2 1 . . .'ilVd8 ! ;
e.g., 22 axb5 cxb4 23 tDa4 !:!.xc 1 24
�xc 1 axb5 25 'iVxb4 bxa4 26 �xb7
ctJc5 ! with the point 27 i..xc5 tDxg5 ! .
After the text-move ( 2 1 ctJh4),
Dautov-Lobron, Essen 1 999 contin­
ued 2 l .. .�d8 22 f4 ctJhf8 23 tDf3 tDb6
(with White's extra space, 23 ... cxb4
24 axb4 tDb6 25 i..d4 tDc4 26 'iVf2! is
not as convincing) 24 bxc5 dxc5 25
'iYf2 ctJfd7 26 :i.f1 ? (White's appar­
ently strong central pawns are re­
strained by nearly every black piece)
26 . . .'iYe7 27 ctJd l (27 . . . tDxd5 was
threatened) 27 ... c4 28 i..c2 �xa3 29
1 76 THE GAMBIT GUIDE TO THE MODERN BENONI
fS i..g7 30 tLld4 tLleS 3 1 fxg6 fxg6 32
tLle6 tLlbd7, when Black was winning.
Note that in this case, the eS-square for
Black was just as good as e6 for White.
82)
7 i..g7 (D)
..•
w are back in the Knight' s Tour of Chap­
Here Black simply allows White to
achieve his desired set-up with tLlf3,
i..d3 and h3. Because of the tremen­
dous importance of this white system,
which as we have seen can also be ar­
rived at by other move-orders, I will
look at this 'second' repertoire option
in the same kind of detail as the first.
8 h3
This is probably the most accurate
move. Otherwise:
a) 8 tLld2 is mentioned in the note
to White's 8th move in Line A of
Chapter 1 0.
b) 8 .i.d3, aiming for the same po­
sition as the main line, allows 8 . . ..i.g4
9 �a4+ (9 0-0 0-0 1 0 h3 .i.xf3 1 1
�xf3 a6 1 2 a4 transposes to Line
B I l l ; one can also play without ... a6
and a4, which has some advantages)
9 . . . tLlbd7 1 0 tLld2 ( 1 0 eS WIIe7) 10 ...0-0
1 1 0-0 ( 1 1 h3 tLlb6) 1 1 ... tLleS 1 2 i.b 1 ,
Smyslov-Tatai, Las Palmas, and now
12 ... .i.d7 1 3 �3 bS ! - Nunn. Black
could also play 8 . . . 0-0, still intending
. . . i.. g4. Then White's only serious in­
dependent option is 9 tLld2, when an
interesting transposition arises from
9 ...tLlbd7 (9 ...tLla6 with the idea . . .tZ:lc7
is fine) 1 0 tLlc4 tLlb6 1 1 tLle3 and we
ter 4 (Line A) !
Here are some less important inde­
pendent continuations:
c) 8 �a4+ and now:
c 1 ) 8 . . ..i.d7 9 �b3 �c7 transposes
to Line B 1 of Chapter 2.
c2) 8 ...tLlbd7 9 .i.e2 (9 i.f4 O-O!
has the idea that 10 i..xd6?? fails to
1O...tLlb6) 9 ...0-0 10 0-0 l:te8 ! ? ( 1 O...a6)
1 1 .i.f4 tLlb6 1 2 'iYc2 tLlhS (or 1 2...i.g4
1 3 h3 i.xf3 1 4 .i.xf3 a6 having in
mind ...tLlc4-eS or .. :fic7 and . . .tLlbd7-
eS) 1 3 i..gS f6 14 .i.e3 fS I S i..gS ( 1 5
i..bS .u.f8) I S . . .tLlf6 =.
d) 8 .i.bS+ .i.d7 (8 . . .tLlfd7, intend­
ing . . . a6, is also satisfactory, as is
8 ... tLlbd7) and now:
d l ) 9 .i.xd7+ tLlbxd7 1 0 .i.f4 ( 1 0
0-0 0-0 1 1 .i.f4 I:i.e'8 1 2 tLld2 tLle5 =)
1 O .. :�e7 1 1 0-0 0-0 1 2 tLld2 a6 1 3 a4
tLleS =.
d2) 9 .i.e2 0-0 10 tLld2 ( l 0 0-0 bS !)
1 O . . .l:te8 1 1 0-0 bS ! 1 2 tLlxbS tLlxe4 1 3
tLlxe4 l.::txe4 = , in view of 14 tLlxd6
.i.a4 I S tLlxe4 .i.xd l 1 6 lIxd l tLld7,
when Black has better pieces.
d3) 9 a4 (innocuous) 9 ... 0-0 1 0 0-0
.i.g4 = 1 1 :el tLlbd7 12 h3 .i.xf3 1 3
�xf3 and although 1 3 . . .a6 14 .i.n was
fine in A.Fernandes-Vitor, Barreiro
 MODERN MAIN LINE
2000, 1 3 . . .lLJe5 ! forces the queen to a
bad square; e.g., 14 'iVdl ( 1 4 �e2 a6)
14 . . .c4 ! .
d4) 9 �b3 0-0 1 0 0-0 ( 1 0 i.xd7
�xd7 is equal) 1 O. . .l:.e8 1 1 l:.e l a6 ! ?
( l l . ..i.xb5 1 2 'ii'xb5 'iWc7 =) 1 2 i.xd7
tiJbxd7 and now, rather than 1 3 a4
'Wic7 = Tregubov-A.lvanov, Russian
Cht (St Petersburg) 1 999, maybe 1 3
�xb7 i s better.
8 0-0 (D)
•••
w I 1 lLJxe4 f5 Stohl; 1 0 e5 lLJe8 Psakhis)
9 i.d3
Other moves are rare:
a) 9 i.f4? l:te8 (or 9 . . . a6 1 0 a4
l:te8) 1 0 i.d3 ( 1 0 lLJd2 a6 1 1 a4 lLJxe4 !
1 2 lLJcxe4 f5) 1 O...lLJxe4 1 1 lLJxe4 f5
1 2 lLJfg5 fxe4 1 3 lLJxe4 i.f5 14 f3 c4!
1 5 i.c2 'iia5+ ! ? ( 1 5 ... i.xe4 ! 16 fxe4
i.xb2 1 7 .l:.b 1 i.c3+ +) 1 6 'it'f1 i.xe4
1 7 i.xe4 i.xb2 1 8 l:.bl . Now, rather
than 1 8 . . . c3, as played in Henderson­
Kinsman, British League (4NCL)
1999/00, 1 8 ...i.e5 might be a bit better;
e.g., 19 i.xe5 1he5 20 l'hb7 lLJa6! 2 1
Wk'bl l:.ae8 22 l:txa7 'iVd2 ! .
b ) 9 i.e3 is inflexible but surpris­
ingly popular, so it deserves some
1 77
attention. Here 9 ...b5 ! ? is sometimes
played, although Black has a number
of other instructive responses:
b l ) 9 ...l:te8 10 lLJd2 a6 (or 1O ...i.d7
1 1 a4 lLJa6 12 i.e2 lLJb4 1 3 0-0 b6 14
i.f3 i.c8 1 5 lLJc4 i.a6 1 6 b3 lLJd7 1 7
l:tc l lLJe5 + Piket-Topalov, Monte
Carlo Amber rpd 1 998) 1 1 a4 lLJxe4 !
1 2 lLJcxe4 f5 1 3 lLJxd6 ( 1 3 lLJxc5 f4)
1 3 ...�xd6 14 lLJc4 �f6 1 5 i.e2 f4 ! 1 6
i.xc5 lLJd7 1 7 i.b4 f3 ! with a terrific
attack, Barria-Minzer, Torre Blanca
1 998.
b2) 9 ... 'iWe7 10 lLJd2 (10 �c2 lLJxe4!
1 O. . . lLJxe4 ! I I lLJdxe4 f5 1 2 lLJxd6
'ii'xd6 1 3 i.e2, Dautov-Psakhis, Dres­
den Z 1 998, and now, among other
good solutions, simply 1 3 . . . a6 14 a4
lLJd7 1 5 0-0 b6 (or 1 5 . . .f4) is fine.
b3) 9 . . .lLJbd7 should also suffice:
10 lLJd2 a6 1 1 a4 ( 1 1 lLJc4 lLJe8 12 a4
lLJe5 =) 1 l . . .l:.b8 1 2 i.d3 ( 1 2 i.e2
lLJe8 1 3 0-0 lLJc7 14 lLJc4 �e7 1 5 i.f4
lLJe5 =) 1 2 ... lLJe5 1 3 i.c2 b5 14 axb5
axb5 1 5 0-0 l:te8 1 6 f4 lLJc4 1 7 lLJxc4
bxc4 =.
We now return to 9 i.d3 (D):
B
1 78 THE GAMBIT GUIDE TO THE MODERN BENONI
9 ttJh5!?
.•.
A somewhat unusual but intriguing
move. I should mention that the alter­
native 9 . . . bS can lead by force, after 1 0
i.xbS ttJxe4 I I ttJxe4 'ifaS+ 1 2 ttJfd2,
to an ending which has been worked
out to past the 20th move and gives ab­
solutely no winning chances for Black
(sometimes White finds minor im­
provements which seem to increase
his own winning chances, and Black
in response finds ways to draw). Such
a move might be acceptable at the
highest levels (in order to draw with
Black), but it would hardly be appro­
priate as a repertoire choice. Black
also has to cope with the enormously
complicated 10 ttJxbS, which at the
moment seems OK for him, but theory
is unsettled.
For the record, I also worked for
many days on the move 9 . . .i.d7, which
theory gives as equal or unclear. Un­
fortunately, that same theory is easy to
improve upon, and I came out con­
vinced that 9 . . . i.d7 gives White a con­
siderable advantage with correct play.
Thus the system before you, which I
think is underrated.
Regarding 9 . . . ttJhS and 1 O. . .ttJbd7
(or 9 . . .ttJbd7 and 1 O. . .ttJhS - see the
next note), I believe that it is particu­
larly important to play these moves (in
either order) without the preliminary
. . . a6 and a4, and also without a prelim­
inary . . .:Le8. The interpolation of ... a6
and a4 can be unfavourable for Black
in several lines, and it can always be
played later if necessary.
Black can normally get to the de­
sired position by playing 9 . . . ttJbd7
first, intending 1 0 0-0 ttJhS, and this
may even be preferable (see the note
to White's 1 0th move). A rare alterna­
tive after 9 ...ttJbd7 is 10 i.f4 ! ? (after
10 ttJbS ttJxe4 !? { or 1O ...c4 followed by
... ttJcS } I I i.xe4 'i'aS+ 1 2 ttJc3 l:.e8
one prefers Black) 1 O. . ..l:.e8 ( 1 O. . .'i1i'e7
also has its points: 1 1 0-0 { I I ttJb5
ttJe8 } 1 1 . . .ttJhS, and now 1 2 i.h2
ttJeS, 12 i.e3 a6 13 a4 :Le8 14 'ii'd2
ttJeS, or 1 2 i.gS ! i.f6! 1 3 .th6 :Le8,
all with perfectly playable positions)
1 1 0-0 ( 1 1 .txd6 Vib6 12 .th2 �xb2
13 llc l c4 14 i.b l ttJg4! +) l l ...c4 1 2
.tc2 ttJcS 1 3 ttJd2 ( 1 3 eS dxeS 1 4
i.xeS i.fS ! I S i.xfS gxfS intending
. . . ttJd3) 1 3 . . .b6 ! ? 14 ttJxc4 ttJcxe4 I S
ttJxe4 ttJxe4 1 6 .:te l ( 1 6 i.a4 i.d7 1 7
i.xd7 'ii'xd7 1 8 f3 bS =) 1 6 . . .i.a6 with
a level game.
10 0-0
White can also try to make progress
by saving the tempo needed for cas­
tling. This probably yields no advan­
tage, but the reader should note that
using the move-order 9 . . .ttJbd7 1 0 0-0
ttJhS would avoid the following lines:
a) 10 g4 weakens the kingside in
return for dubious benefits. Normally
. . .hS follows at some point, as we will
also see in later lines; e.g., 1O . . .ttJf6 1 1
.tf4 hS ( 1 l . ..�e7 =) 1 2 gxhS ( 1 2 ttJd2
hxg4 1 3 ttJc4 gxh3 14 ttJxd6 .tg4 with
the idea I S f3 ttJhS; 1 2 gS ttJe8 practi­
cally forces White's king to the vul­
nerable queenside, and usually makes
a later .. .f6 or . . .fS effective) 1 2...ttJxhS
1 3 .tgS .tf6 14 i.e3 ttJd7 I S �d2
.:te8 =.
b) 10 .tgS is the only serious al­
ternative: 1 0 . . .i.f6 1 1 i.e3 ( 1 1 i.h6
 MODERN MAIN LINE
i.g7 !? { 1 l ....:i.e8 is also fully playable }
12 i.xg7 �xg7 1 3 0-0 �f6 ! = with the
ideas of . . .lLlbd7 and . . . lLlf4) 1 l . ..lLld7
12 i.e2 ( 1 2 g4?! lLlg7 1 3 gS i.e7 1 4
h4, Antwerpen-Hasselt, COIT. 1 99 1 ,
and now 1 4 ...fS { Schneider } or just
14 . . .f6 !) 1 2 . . Jle8 1 3 lLld2 lLlg7 1 4 0-0
(14 lLlc4 i.xc3+ IS bxc3 lLlf6 1 6 f3
tbfhS = with ideas like . . .lLlg3, ...bS
and .. .fS) 14 . . .i.xc3 ! ? I S bxc3 fS ! . I
think that Black is OK here, but if you
don't like it, consider 9 . . .lLlbd7 first.
10 tbd7 (D)
..•
Our key position. It's worth men­
tioning that much of the theory of this
type of position derives from the lines
in which . . . a6 and a4 are interpolated.
In the appropriate places, I will rec­
ommend transposition to those lines.
What is Black up to? For one thing,
the move . . .lLlhS prevents i.f4, and
that is no small matter, since i.f4 is
played in almost every Modem Main
subsystem. Secondly, the move ... tbeS
becomes available, either before or af­
ter moves like . . .l:le8. If one looks at
the theory of other defences to the
1 79
Modem Main Line, one sees that Black
has a lot of difficulty clearing the first
rank, in part because the d7-knight
gets in the way. In this case, . . .lLleS
both hits d3 and frees the c8-bishop.
After lLlxeS and . . .i.xeS, Black begins
to look towards the kingside, with
. . .'iVh4 being particularly appropriate
since h3 has been played. White can
respond to ... lLleS with i.e2, but we
will see that this opens up other tacti­
cal possibilities. Finally, the move g4
is usually quite weakening when com­
bined with White's rather slow set-up
with h3, i.d3, etc. As a rule, White
should delay this tactical thrust until
he has developed more pieces.
Due to these considerations, White
tends to move his queen's bishop at
this point, both to develop and to dis­
turb Black's plans. Alternatively, he
can prepare the retreat i.f1 by playing
l:lel .
B21 : 1 1 i.g5 1 79
B22: 11 i.e3 1 82
B23: l1 l:tel 1 83
Or:
a) 1 1 tbbs lLle5 =; e.g., 1 2 g4 lLlxd3
1 3 'i'ixd3 a6! .
b) 1 1 g4 i s playable but somewhat
weakening: l l . . .lLlhf6 1 2 i.f4 'i'ie7 1 3
:Le i ! ( 1 3 lLlbS tbe8) 1 3 . . . hS ! ? (or
1 3 . . .lLle8 ! =) 14 eS lLlxeS I S i.xeS
dxeS 16 lLlxeS 'i'id6 ( 1 6 ...'i'id8 17 'i'if3
hxg4 1 8 hxg4 lLld7) 1 7 lLlbS ! ? 'i'id8 =.
82 1 )
11 i.g5
Probably the most aggressive move,
and the traditional response to . . .lLlhS.
180 THE GAMBIT GUIDE TO THE MODERN BENONI
White tries to disrupt Black' s devel­
opment by forcing weaknesses or
awkward piece placement.
1l i.f6 (D)
••.
1 1 . . :ii'b6 ! ? also merits consider­
ation; e.g., 1 2 'ii'd2 lLleS 1 3 lLlxeS
.i.xeS 14 i.e3, De Boer-Faase, Neth­
erlands 1994, and here 14 . . . i.d7 I S f4
i.d4 ! ? 1 6 .i.xd4 cxd4 1 7 lLle2 1:tae8
intends . . .fS, and 1 8 fS gxfS 1 9 exfS
lIe3 20 f6 'it>h8 is unclear.
12 .i.h6
Or:
a) 12 .i.xf6 'ii'xf6 = with the ideas
. . .lLlf4, . . . a6 and . . .lLleS; e.g., 1 3 g3
l:te8 14 l:tel ( 1 4 lLlbS lLlb6 !) 14 . . . a6 I S
a4 lLleS 1 6 lLlxeS l:.xeS with an equal,
double-edged game.
b) 12 .i.e3 is also important:
b l ) 12 . . .lLleS ! ? and here:
b 1 1 ) 1 3 .i.e2 lLlxf3+ 14 i.xf3 lLlg7 !
I S 'ii'd2 i.eS 1 6 i.f4, Sokolin-Sandic,
Pula 1 989, and now Kapengut sug­
gests simply 16 . . .l:.e8 with the idea
. . .fS; this is equal. 16 ... i.xf4 17 'ii'xf4
gS ! ? 1 8 'ii'd2 fS, with equality, is also
of note.
b 1 2) 1 3 lLlxeS i.xeS 14 f4 i.g7 I S
'ii'f3 ! i s more dangerous; compare the
less threatening positions with the
queen on d2.
b2) 1 2. . .l:te8 and now:
b2 1 ) 1 3 g4 lLlg7 14 gS .i.e7 (or
14 ....i.eS ! ?) IS h4 f6 16 'itg2 l:tf8 1 7
l:th l lLleS ! ( 1 7 . . .lLlhS ; 1 7 . . . fxgS 1 8
hxgS lLlhS = ; e.g., 1 9 'ii'd2 lLles 20 i.e2
lLlg4) 1 8 lLlxeS fxeS 1 9 l:tb I lLlh5 with
...lLlf4+ next.
b22) 1 3 :bl a6 (or 1 3 . . .lLle5 1 4
.i.e2 lLlxf3+ 1 5 i.xf3 lLlg7) 14 a4 lLle5
( 1 4. . .b6 1 5 'ii'd2 lLleS 16 lLlxe5 .i.xe5
17 b4 l:.a7 ! ? { 1 7 . . . i.d7 } 1 8 bxc5 bxc5
1 9 lLle2 i.g7 =) 1 5 i.e2 lLlxf3+ 1 6
.i.xf3 lLlg7 1 7 i.f4 .i.e5 1 8 .i.xeS l:txe5
1 9 b4 b6 20 bxcS bxcS (or 20. . .dxc5
intending . . . lLle8-d6) 2 1 .i.e2 f5 =.
b23) 13 'ii'd2 lLleS 1 4 .i.e2 ( 14
lLlxe5 i.xeS is equal, since this time
15 f4? fails to IS ....i.xc3 16 'ii'xc3 lLlg3,
and the insertion of 'it'd2 and . . .l:.e8
helps Black in any case) 14 . . .lLlxf3+
1 5 i.xf3 lLlg7 and Black intends
. . . .i.e5 and . . .fS ; for example, 16 :fel
( 1 6 'ii'c2 i.gS) 1 6. . ..i.eS 1 7 .i.g5 f6 1 8
.i.h6 a6 1 9 a4 b6 intending . . .l:.a7-f7
and/or . . . fS.
b24) 1 3 lLld2 a6 14 a4 i.d4 trans­
poses into Krasenkow-Oliwa, Polish
Ch (Brzeg Dolny) 1 996: IS i.xd4 cxd4
1 6 lLle2 lLlcs 1 7 l:ta3 i.d7 1 8 i.c2 d3
1 9 i.xd3 .i.xa4 20 l:txa4 lLlxd3 2 1
l:.d4, and now Black should have
played 2 l ...lLlhf4 ! ; e.g., 22 lLlxf4 (22
lLlc3 lLlxb2 + 23 'it'b3? 'iWg5 24 g3
:ac8 ! ) 22 . . .lLlxf4 23 'it'f3 �gS 24
lLlc4 l:.ad8 2S g3?! 'it'f6.
A theme to note in these lines is
how the apparently poor position of
 MODERN MAIN LINE
the knight on g7 is compensated by its
ability to enforce . . .fS or . . . hS. This
theme arises in several Benoni sys­
tems.
We now return to 1 2 �h6 (D):
B try 20. . ..l:.b8 ! ? 2 1 :adl b4 22 axb4
12 .. J�e8!?
The ambitious move, but it should
be said that 1 2 . . ..i.g7 is also a reason­
able continuation: 1 3 .i.xg7 ( 1 3 'iWd2
i.xh6 ! ? { 1 3 . . .:e8 } 1 4 'iWxh6 'iWf6 { '='
Kapengut, but 14 ... f6 ! should be con­
sidered as well, restricting White's
bad bishop } I S .i.bS ! 'iWg7 1 6 'iWxg7+
<Jitxg7 1 7 lLld2 .l:.d8 1 8 f4 a6 1 9 �xd7
i.xd7 20 g4 lLlf6 2 1 :ac l t Heigl­
Beyer, corr. 1 995) 1 3 . . .�xg7 (Kapen­
gut suggests 1 3 . . .lLlxg7 1 4 'iWd2 'iWf6,
but this seems a worse version of the
given line, since I S 'iWh6 is available,
among other ideas) 14 'iWd2 'iWf6 ! with
the idea of controlling f4; if I S lLle2,
then I S . . . .l:.e8 is sensible, with ideas
like . . .bS and . . .lLleS.
13 'ifd2
1 3 g4 �g7 ! ? ( l 3 . . . lLlg7 is the alter­
native, and might contain some new
ideas; a possible continuation would
181
be 14 .i.f4 hS I S gS { I S .i.xd6 hxg4 1 6
hxg4 'ifb6 = } I S . . . �eS 1 6 lLlxeS
lLlxeS 1 7 �g2 f6 =) 14 .i.gS lLlhf6 ! ?
( l 4. . . �f6! =) I S l:te l a6 1 6 'iWd2 ( 1 6
a4 b6) 1 6 . . .b S 1 7 'iWf4 'iWe7 1 8 a3 ! t hS
19 'ifg3 hxg4 20 hxg4 and here, instead
of 20. . .'iffS as in Atalik-C.Hansen,
Thessaloniki OL 1 988, Black might
.l:.xb4 23 .l:.d2! (23 .i.f4 lLleS) 23 . . .c4 =,
although one feels that either 1 3 . . .lLlg7
or 14 . . .�f6 is a better way to equality.
13 a6 14 g4
•••
White reasons that he has enough
forces out to justify this move. Other­
wise, 14 a4 lLleS ( 1 4. . .b6 ! ?) I S lLlxeS
.i.xeS gives Black a reliable position.
14 lLlg7
.••
OK, what's the knight doing on this
awful square? But equally, what's
White's pawn doing on g4? As dis­
cussed earlier, Black can reorganize
his pieces by bringing a knight to cS or
eS, and it will be hard for White to re­
strain both . . .fS or . . . hS. See also sev­
eral examples of a knight on g7 in note
'b' to White's 1 2th move.
15 a4 'ii'c 7!?
I S . . .b6 is a slower but solid move,
intending . . . lLleS, with . . ..l:.a7 when ap­
propriate to defend the kingside, e.g.,
after ... fS.
16 'it>g2 c4 17 .i.c2 .l:.b8 18 lLld4
lLlc5 19 g5 'iWd7! 20 l:thl .i.d8
Now the bishop will be active on
as, and White's pieces are hardly well
positioned for attack.
21 �xg7 'it>xg7 22 f4 �a5 23 :ael
�g8
Now, instead of 24 'iWdl ? 'ifd8 ! 2S h4
'ifb6 += Zaichik-Romanishin, Tbilisi
182 THE GAMBIT GUIDE TO THE MODERN BENONI
1988, Zaichik recommends 24 fS with
the idea h4-hS . Unfortunately, Schnei­
der points out that this allows simply
24 ... ttJxa4 ! threatening . . . ttJxb2; for
example, 2S ':'al .i.xc3 26 bxc3 ttJcS
27 �f4 "ike7 with advantage.
822)
11 .i.e3
Instead of harassing Black, White
counts upon direct development. This
bears comparison with 1 1 .i.gS .i.f6
1 2 �e3 (note 'b' to White's 1 2th move
in Line B2 1 ) ; the interpolation of
...�f6 does not clearly help either side.
11 a6 (D)
.•.
l 1 . . .ttJeS 1 2 ttJxeS �xeS 1 3 f4
�g7 1 4 'iWo b6 I S ':'ae1 ( 1 S g4 ttJf6
1 6 eS dxeS 1 7 fxeS ttJd7 1 8 d6 .:tb8 'f)
I S . . . .i.b7 is maybe a bit better for
White, but that's hard to prove.
w preparing a second-rank transfer by
12 a4
This is a position with some theory
behind it, since ... a6 and a4 have been
interpolated. Black seems to have more
than one way to achieve a dynamic
balance.
The alternative 1 2 g4 ! ? is weaken­
ing, as usual, but it also gains time to
contest eS. Some hopefully represen­
tative analysis follows, so as to famil­
iarize ourselves with the themes:
12 ...ttJhf6 13 .i.f4 ike7 14 .:tel ttJeS (we
have encountered the idea 1 4 ... hS ! ? I S
eS dxeS 1 6 ttJxeS ttJxeS 1 7 l:.xe5
ikd6 ! ? before) 1 5 ttJxeS dxeS 16 �e3
b6 1 7 iko hS ! 1 8 gS ( 1 8 gxhS ttJxhS 1 9
':'ad l ikh4 20 'it'h2 l:.a7 =) 1 8. . .ttJh7
1 9 ikg3 f6 ! (the standard break we so
often see in lines with gS) 20 gxf6
':'xf6, and White's position is under
some pressure; e.g., 2 1 l:.ad l 'it'h8 22
f4? (22 a3 gS ! 23 �xgS ttJxgS 24
'ii'xgS .i.xh3 with a very promising at­
tack after ... �g4 and ... .i.h6) 22 ...exf4
23 i.xf4 Ihf4! 24 'ii'xf4 .i.eS 2S ikfl
(2S ikh6 ikh4 -+) 2S ... ttJgS =t.
12...b6!?
This is an attempt to keep more
play in the position, reinforcing cS
against eS ideas and in some cases
....:.a7-e7. Black can also choose:
a) 12 ....:.b8 ! ? 1 3 "ikd2 ( 1 3 �c2! ?, to
answer 1 3 ...ikaS with 1 4 ttJd2, is met
by 1 3 . . . .:.e8 intending . . .ttJeS, when
the queen isn't particularly well-placed
on c2) 1 3 ....:.e8 1 4 ':'abl ikaS yielded
mutual chances in G.Georgadze­
B .Stein, Gausdal 1 992.
b) 1 2....:.e8 1 3 ':'el ttJeS ! ? ( 1 3 . . .b6
is solid) can lead to complex and forc­
ing play: 1 4 i.e2 ttJxO+ I S .i.xf3 ttJf6 .
1 6 i.f4 ttJd7 ! 1 7 .i.xd6 ! 'ii'b6 1 8 eS .
ttJxeS 1 9 i.xeS ':'xeS ! 20 l:.xeS i.xeS
21 ike2 f6 22 as ikc7 23 ikc4 ! (other­
wise Black's bishops get out) 23 ...�g7
24 ttJe4 .i.fS ! 2S d6 (2S ttJxcS ':'c8 26
 MODERN MAIN LINE
ttJe6+ .i.xe6 27 'ii'xc7+ .:txc7 28 dxe6
�f8 and 2S 'iVxcs .i.xe4 26 'iVxc7+
j.xc7 27 .i.xe4 .i.d6 should be drawn
easily) 2S . . . i.xd6 26 tDxd6 'iVxd6 27
.i.xb7 l:.a7 28 i.f3 .i.d3 29 l:.d l l:.d7 =
P.Cramling-Yudasin, Dos Hermanas
1 992.
13 'iVd2 tDe5! 14 i.e2
14 tDxeS i.xeS is less critical: I S f4
( 1 S l:.fel l:.e8 1 6 1bbl 'it'h4 ! ? 1 7 .i.gS
.i.f4 1 8 i.xh4 i.xd2 19 .l:.ed 1 i.h6 20
.i.e2 tDf4 2 1 i.fl i.g7 =) I S ....i.g7
(considering ideas like ...'iVh4) 16 g4 ! ?
ttJf6 (with the threat . . . hS) 1 7 'iWg2
(Quinn) 17 . . .l:.a7 ( l 7 . . .l:.e8 1 8 :ae l
h6 ! ?) 1 8 gS tDhS 1 9 .i.e2 ( 1 9 l:.ael fS)
1 9 . . . fS ! 20 i.xhS gxhS and the a7-
rook ensures satisfactory counterplay
for Black.
14 f5!? 15 tDxe5 i.xe5 16 exf5!
.•.
Or:
a) 1 6 f4? ! i.xc3 1 7 bxc3 tDg3 ! .
b) 1 6 i.xhS f4 ! ? 1 7 .i.f3 'ii'gS 1 8
..t>h 1 'ii'h4 should be at least equal,
since 1 9 i.xcs bxcS gives Black a
ready-made attack by . . .gS and ...hS or
... .:tg7 and . . . g4.
16 ....i.xf5 (D)
183
17 g4?
This just doesn't work out tactically
with so many of Black's pieces aimed
at the kingside. Better is 1 7 i.gS tDf6
( 1 7 . . . i.f6 is perfectly reasonable) 1 8
g4 ( 1 8 l:.fel 'i'c8 ! ? 1 9 .i.fl { 1 9 i.c4
i.xh3 ! } 19 ... l:.a7 20 g4 i.xg4 2 1 hxg4
'iWxg4+ 22 .i.g2 'iWhS with an obscure
but dangerous attack) 1 8 . . . .i.d7 ! ? (or
1 8 . . ..i.c8 1 9 .i.c4 l:.a7 20 l:tae1 l:te7 = )
1 9 i.c4 'iWc8 (threatening . . ..i.xg4) 20
f3 bS ! 2 1 axbS axbS 22 tDxbS .i.xbS
23 .i.xbS ':'b8 24 ':'as ! ? c4 with good
compensation.
17...'iVh4 18 ..t>g2 .i.f4!
A very nice touch. White is in ma­
jor trouble now; for example, 1 9 l:th 1
( 1 9 gxfS .i.xe3 20 'iWxe3 { 20 fxe3
'iVg3+ 2 1 'its'h1 'iWxh3+ 22 �gl tDg3
with an attack } 20. . .l:.ae8 -+; 1 9 Wgl
'iWxh3 20 .i.xf4 i.e4 -+) 19 ... .i.xe3 20
'iWxe3 l:tae8 2 1 'iWd2 l:.xe2 ! and Black
wins. The play following 14 ... fS was
expanded from analysis by Mark
Quinn.
823)
11 l:.el (D)
184 THE GAMBIT GUIDE TO THE MODERN BENONI
1l lLleS!?
•••
This move leads to fascinating
complications. Other ideas:
a) The main alternative is 1 l . ...l:.e8,
which resembles note 'b' to Black's
1 2th move in Line B22, although
.. Jle8 was probably better timed in that
case: 1 2 i.e3 ( 1 2 i.f1 a6 ! { 1 2 . . . lLleS
1 3 lLlh2 ! t is more favourable for
White } 1 3 a4 lLleS 1 4 lLlh2 �h4 ! is
unclear) 12 ...lLleS 13 lLlxeS ! ( 1 3 i.e2
lLlxf3+ 14 i.xf3 lLlf6, with the idea
that IS i.f4 is met by I s ...lLld7 ! 1 6
i.xd6 'iib6) 1 3 . . .i.xeS 1 4 'iid2 t.
b) For the record, it is instructive to
compare a line like 1 1 ...a6 12 a4 lLleS
1 3 i.e2 lLlxf3+ 1 4 i.xf3 'iih4 I S
i.xhS gxhS 1 6 'iid2 fS 1 7 l:ta3 ! f4 1 8
lLle2 f3 1 9 lLlf4 ! with our main line be­
low.
12 i.e2!
1 2 lLlxeS i.xeS 1 3 i.h6 .l:.e8 1 4
'iid2 a6 I S a4 'iif6 (or I S . . .'iih4 ! ? 1 6
i.gS i.f4 1 7 i.xh4 i.xd2 1 8 l:ted 1
i.h6 = ) 1 6 i.gS i.f4 1 7 'iixf4 ( 1 7
i.xf6 i.xd2 1 8 l:tedl i.f4 ! 1 9 i.h4 gS
+) 1 7 . . .'iixf4 1 8 i.xf4 lLlxf4 =.
12 lLlxf3+ 13 i.xf3 'iih4! (D)
.•.
w i.xd6 ! ! = , while 1 7. . .i.h2+ ! ? may re­
This move is dubious with ... a6 and
a4 in, mainly because l:ta3 ! is a good
move for White at one point, and also
because as is effective in some lines -
compare note 'b' to Black's 1 1 th move.
But here ...'iih4 gives good counter­
play in return for Black's shattered
pawns.
14 i.xhS
14 eS? ! i.xeS I S l:te4 'iif6 1 6 i.xhS
gxhS 17 'iixhS i.fS almost certainly
gives Black the advantage with his en­
ergetic bishop-pair.
14 gxhS
•..
In return for the shattered pawns,
Black has two active bishops, the
queenside majority and attacking ideas
involving the open g-file. We come to
an important decision:
IS 'iid2
A very challenging alternative that
was not available in the ... a6/a4 lines
mentioned above is I S lLlbS ! ?, al­
though after intriguing complications,
IS . . . i.eS appears both sound and
promising for Black; e.g., 1 6 i.h6 ( 1 6
lLlc7 l:tb8 1 7 i.h6? i.xh3 ! ; 1 6 'iic2
can be met by 16 . . .i.d7 or 16 ...�h8 ! =+=
with ideas like . . . i.xh3 and . . . l:tg8)
16 ... i.xh3 ! 17 �b3 ! ( 1 7 i.xf8 can be
met by 1 7 . . . i.d7 ! 1 8 f4 ! i.xf4 1 9
tain winning chances amidst the chaos;
17 gxh3 'iixh3 1 8 f4 �h8 ! 1 9 i.xf8
l:txf8 + intending ... .l:.g8) 17 . . .i.d7 ! 1 8
g3 ! ( 1 8 i.xf8 'iih2+ 1 9 �f1 l:txf8 +;
White can hardly defend) 1 8 ... 'iif6
( 1 8 ... c4 ! ?) 1 9 i.xf8 l:txf8 and Black
has more than enough counterplay,
since the exchange for a pawn doesn't
mean much with ... i.xbS, ... h4 and
 MODERN MAIN LINE
...�hS coming. White can't even bail
out with 20 f4 .ixb5 ! 2 1 fxe5 'ii'xe5 22
"iYxb5, since 22 .. :ii'x g3+ 23 �h 1
"iYh3+ 24 �g 1 WhS 25 �f2 %lgS
leaves Black well on top. As always,
the Benoni is an exercise in sustained
initiative.
15 f5! 16 'ii'g5!
•••
16 'ii'f4 'ii'f6 17 ltJb5 fxe4 IS 'ii'xf6
lhf6 = ; I like Black's bishops.
16 :iWxg5 17 .ixg5 fxe4 18 .ie7
.•
I S ltJxe4 is harmless in view of
1 8 . . . .ixb2: 1 9 :ad 1 .if5 20 ltJxd6
i.c2 = or 1 9 %lab1 ? ! .id4 ! .
I think White's play has been opti­
mal up to this point, but his slight ad­
vantage is meaningless in view of the
coming opposite-coloured bishops;
e.g., l S ...:eS ( l S ...lIf7 19 .ixd6 b6 is
also possible) 19 .ixd6 ( 1 9 l::txe4 .if5
20 lIf4 :xe7 2 1 lIxf5 :fS is equal)
19 ... .ixc3 20 bxc3 :dS 2 1 .if4 (2 1
i.xc5 lIxd5 22 .id4 .if5 =) 2 l ...:xd5
22 lIxe4 .id7 23 c4 (23 :ae1 :fS = )
185
23 ...:d4 24 %lxd4 cxd4 25 :d1 lIcs
26 :xd4 .ie6 and the draw is obvious.
This whole variation with . . .ltJh5
and ...ltJbd7 has been seriously under­
estimated, in my opinion, and offers
Black better chances than lines such
as 9 . . .%leS, 9 . . .ltJa6, 9 . . . .id7, 9 . . . c4,
and the corresponding lines following
9 . . . a6 10 a4. I won't pretend that I
don't have some intuitive doubts about
this system achieving absolute and in­
controvertible equality (although it
may do so). On the other hand, what­
ever advantage White has is probably
on the scale of his slight advantage in
any other opening. For the practical
Benoni player, this method of play of­
fers the sort of double-edged fight
most players want, without inordinate
risk for Black.
The Modem Main Line is still a for­
midable weapon, but not one that
should frighten you into looking for
another opening !
1 0 Classica l M a i n L i ne
1 d4 tLlf6 2 c4 c5 3 d5 e6 4 tLlc3 exd5 5
cxd5 d6 6 tLlf3
Move-orders leading to the Classi­
cal Main Line are very tricky and full
of implications. 6 e4 g6 7 tLlf3 allows
7 . . . a6, as discussed in Chapter 9, so 6
tLlf3 is the most flexible way to
achieve an early tLld2 without allow­
ing either . . .iLg4 or . . .bS . Note that 6
e4 g6 7 iLe2 iLg7 8 tLlf3 is a move­
order briefly discussed in Chapter 1
(note 'f5 ' to White's 7th move in Line
D2).
6 g6
.•.
Here White has a choice:
A: 7 e4 1 86
B: 7 tLld2 1 99
This division is artificial, since the
main lines of both sections converge
by move 10. I have arranged the mate­
rial so as to emphasize move-order
issues, while retaining the independ­
ence of each section, so that Line A
will cover . . .tLlbd7 lines and Line B
will examine . . .tLla6 lines.
A)
7 e4
This is the traditional way to get to
the Classical Main line, but as men­
tioned above, Black can now deviate
by 7 . . . a6. The idea is that after 8 a4,
Black can play 8 . . . iLg4 and reach a
variation we dealt with in the previous
chapter. That is held in high regard for
Black, so much so that some consider
it a good reason not to play these
Classical lines with 7 e4 at all. But of
course, Black may not like other sub­
variations which may arise from this
move-order, such as 7 . . . a6 8 h3 bS 9
iLd3 (again, see Chapter 9); for that
reason or others, he may bypass 7 . . .a6
and play directly into this chapter with
7 . . .iLg7. Note, by the way, that 7 e4 a6
8 iLe2?! is poor for White, since 8 . . .b5
already threatens . . .b4.
7 iLg7 8 iLe2
.••
8 'iVa4+ and 8 iLd3 are dealt with in
the note to White's 8th move in Line
B2 of Chapter 9 (lines 'c' and 'b' re­
spectively). If White still wants to pre­
vent the possibility of . . . iLg4, he can
play 8 tLld2 now, although this is
slightly less flexible than 7 tLld2. In
the latter case, Black is forced to com­
mit to his favourite defence against the
Knight's Tour Variation (Chapter 4),
and White can do without e4 if he so
chooses. Anyway, after 8 tLld2 0-0 9
iLe2 we would rejoin this chapter.
8 0-0 9 0-0 (D)
•••
This is the start of the Classical
Main Line, in which White puts his
pieces on 'natural' squares and hopes
to restrict Black's possibilities before
mobilizing for attack. The Classical
Main Line, formerly the principal line
 CLASSICAL MAIN LINE
of the Benoni, is no longer as popular
as it once was, having been superseded
by the Modem Main Line (Chapter 9).
In part, this is because fairly standard
remedies have been found for Black,
and in part because the Modem Main
Line is so difficult to meet. There are
nevertheless quite a few strong play­
ers who play the traditional system,
and no serious Benoni player can af­
ford to be unprepared for it.
9 .lte8!?
•.
A move that works in this position,
but is needed even more versus move­
orders in which White plays an early
ctJd2 and then e4. A few more fine
points:
a) 9 ... ctJa6 appears to save a tempo
( . . . .l:!.e8) on the ... ctJa6-c7 lines which
are presented in Line B. But White can
take advantage of the omission and
forego 1 0 ctJd2 in favour of 1 0 SLf4 ! .
Then theory gives White the nod.
b) On the other hand, if you got
this exact position (i.e., where White
has decided not to play ctJd2 earlier),
you would probably be happy to play
our first choice 9 . . . a6! 1O a4 SLg4! (D).
187
Then:
b l ) 1 1 h3 i.xf3 just wastes a tempo
or commits White to an h3 he may not
want to play; Black was going to ex­
change on f3 anyway.
b2) 1 1 i.f4 SLxf3 1 2 SLxf3 brings
us to a key line from the suggested
system of Chapter 9 (Line B I 1 2). This
was the move-order that originally
drove White away from the Classical.
Today, most players try to avoid it by 8
ctJd2 or 9 ctJd2.
b3) White can still deviate by play­
ing 1 1 ctJd2 SLxe2 1 2 'iWxe2, which
unfortunately is very complicated:
12 . . .ctJbd7 1 3 ctJc4 ( 1 3 f4 l:te8; 1 3 a5
'iWe7 1 4 l:tel l:tae8) 1 3 . . .ctJb6 1 4 ctJe3
(14 ctJa3 l:te8 15 'iWc2 ctJh5 ! ? 16 a5 ctJd7
1 7 ctJc4 ctJe5 1 8 ctJb6 l:tb8 = Lputian­
Suba, Debrecen Echt 1 992; 14 a5
ctJxc4 1 5 'iWxc4 ctJd7 ! =) 14 ...'iWe7 1 5
a5 ltJbd7 1 6 ltJc4 ltJe5 1 7 ltJb6 l:tae8
1 8 i.g5 h6 1 9 i.h4 and now:
b3 1 ) After 1 9 ...'ifc7 ! ? 20 f4 ltJed7
Black played 2 1 ltJc4?! (21 ltJxd7 ! is
more awkward) in M.Kovacs-Nicev­
ski, DeCin 1978, and now Black should
188 THE GAMBIT GUIDE TO THE MODERN BENONI
have played Kovacs's own proposal
of 2 1 ... tiJxe4! 22 tiJxe4 f5 23 tiJcxd6
'iWxd6 24 tiJxd6 ':xe2 with at least
equality.
b32) I prefer Kapengut's 1 9 . . .g5 ! ?
20 i..g3 tiJfd7, since .. .f5 will arrive
before any tiJd l -e3 manoeuvre. After
2 1 f4 gxf4 22 tiJxd7 he continues
22. . . tiJxd7? 23 ':xf4, which must fa­
vour White due to his control of f5 .
However, 22 . . .'iWxd7 ! forces matters:
23 lhf4 (23 i..xf4 f5 =) 23 . . .f5 24
:afl tiJg6! 25 :4f3 (25 ':xf5? ':xf5
26 ':xf5 i..xc3 !) 25 ...i.. d4+ 26 <J;>h l
i..xc3 27 bxc3 ':xe4 with a small edge
for Black.
So ultimately, 1 1 tiJd2 is no threat
to Black's move-order of 9 ... a6 1 0 a4
i..g4.
To summarize: our preferred course
against the natural move-order of the
Classical Main given in this section
would be 9 ... a6 1 0 a4 i..g4, when 1 1
i..f4 i..xf3 1 2 i..xf3 is analysed thor­
oughly in Line B 1 1 2 of Chapter 9.
That is a fairly easy and dynamic posi­
tion to play, but it is not available
against an early tiJd2. What follows
after the text-move (9 . . . ':e8) is an-
other system involving . . .tiJbd7 which
could be used against both the move­
order in this section and against early
tiJd2s by White (that is, 7 tiJd2 or 8
tiJd2). Finally, in Line B, I will suggest
a system involving ...tiJa6-ideas, which
is also applicable to the move-order
before us, should White play 1 0 tiJd2
next instead of 1 0 'iWc2 (see the next
note).
We now return to the position after
9 . . ..:te8 (D):
w
10 tiJd2
1 0 'iWc2 is the only other logical
way to defend the e-pawn. One reason
that this isn't seen more often is
1 O ... tiJa6, but it's also worth noting
that 1 O...i.. g4 ! is an excellent alterna­
tive which has consistently equalized
in limited experience. I actually prefer
that move, which one might compare
with similar lines in Chapter 9. Re­
turning to 1 0... tiJa6, White has these
options:
a) 1 1 i..xa6?! bxa6 and . . . ':b8 is
just about always bad for White.
b) 1 1 tiJd2 :b8 =; e.g., 1 2 h3 ( 1 2 f4
tiJc7 1 3 a4? tiJfxd5 ! 14 exd5 i..xc3 1 5
'iWxc3 1:txe2 + ) 1 2. . .tiJc7 1 3 a4? (but
otherwise . . . b5) 1 3 ...tiJfxd5 ! 1 4 exd5
i..xc3 1 5 'iWxc3 ':xe2 1 6 'iVf3 (Schnei­
der) 1 6 ... :e5 ! +.
c) 1 1 ':el i..g4 ( 1 1 ...tiJc7 is a rea­
sonable alternative) 1 2 i..f4 and now:
c l ) 1 2 ....:tc8 is logical, preparing
... c4 and ... tiJb4.
c2) 1 2 . . . c4 ! ? 1 3 i.. xc4 i.. xf3 1 4
gxf3 tiJh5 1 5 i..g3 i..e5 1 6 i..xa6 ( 1 6
i..b5) 1 6. . .bxa6 17 tiJe2 'ii'f6 gave Black
some compensation in Nemet-Gobet,
 CLASSICAL MAIN LINE
Biel 1 983. This is awfully speculative,
however.
c3) A good alternative is l 2 ...ltJhS
1 3 i.gS i.f6 ! ? (Black should investi­
gate 1 3 . . .f6 14 i.e3 fS ! , having in
mind I S ltJgS f4 ! or I S i.gS i.f6) 1 4
.te3 ltJc7 ( 14... c4?! I s lLld2 .Jtxe2 1 6
l:txe2 lLlb4 1 7 'ii'b l gives White a
slight advantage - ECO) IS h3 .Jtxf3
1 6 i.xf3 lLlg7 with the ideas ...bS and
. . . i.eS followed by . . . fS.
d) 1 1 a3 lLlc7 1 2 lIel ( 1 2 i.gS .:tb8
1 3 a4 a6 14 as bS IS axb6 lhb6 =
Fedorov-Maslak, St Petersburg Chi­
gorin mem 2000) 1 2 ...l:.b8 1 3 .i.f4 bS
14 b4 ! ? lLlxe4 ! I S lLlxe4 i.fS 1 6 lLlfd2
lLlxdS = Veksler-Shestoperov, Omsk
1973.
e) 1 1 i.f4 ltJb4 12 'ii'b l lLlhS 1 3
.tgS f6 1 4 i.e3 ( 1 4 i.h4 lLlf4) 1 4. . .fS
IS a3 fxe4 1 6 ltJgS ltJd3 1 7 .JtxhS gxhS
18 lLlgxe4 c4 1 9 'iWd 1 i.fS 20 'ii'f3
�d7 2 1 i.gS, Portisch-Adamski, Raach
Z 1 969, and now there could follow
2 1 .. .lIf8 ! ? 22 'iWxhS i.xe4 23 lLlxe4
i.xb2 ! , which is complex but dynami­
cally equal (Kapengut's "23 . . . ltJxb2
+" has several good answers, espe­
cially Hiares's 24 'iWh4!).
IO lLlbd7 (D)
•••
In the next section (Line B), we will
examine the alternative IO . . .ltJa6 in­
tending . . . lLlc7.
Now we have:
AI: 11 h3 1 90
A2: 11 a4 1 94
1 1 a4 (or transpositions to it) has
been the overwhelming choice of
players for many years. But 1 1 h3
causes unique problems and should be
considered separately.
189
There are several meaningful alter­
natives:
a) 1 1 .:tel lLleS ( 1 l . ..h6 intending
. . .gS is also logical and should be com­
pared with the main lines) 1 2 lLln ( 1 2
f4 lLleg4 1 3 .Jtf3 hS ! 1 4 h3 lLlh7 ! with
an attack, according to Schneider)
12 ... a6 1 3 a4 l:.b8 ( 1 3 . . .hS ! ?) 14 h3 h6
I S ltJg3 gS 16 .i.e3 g4 ! ? with compli­
cations - analysis by Schneider.
b) 1 1 f4 is a well-known variation
of the Four Pawns Attack. In some­
what abridged form, the main line
goes l l . . .c4! 12 'iti>h l ( 1 2 .Jtxc4 lLlcs
13 eS dxeS 14 fxeS lIxeS =) 12 . . .ltJcS 1 3
eS dxeS 1 4 fxeS l:.xeS I S lLlxc4 l:.e8
16 i.gS h6 1 7 .i.h4 lLlce4 ( 1 7 . . . .i.fS ! ?
1 8 d6 'ii'd7 should also be fine for
Black, Gorelov-Petrushin, Aktiubinsk
1 985) 1 8 lLlxe4 ':'xe4 1 9 i.g3 'ii'xdS
20 'iWxds lLlxdS 2 1 i.f3 ':'d4 22 ':'ad l ,
Chandler-Sax, Sarajevo 1 985, and now
Schneider gives 22 . . .lLlb4! 23 ':'xd4
i.xd4 24 ':'d l .i.f6 2S a3 lLlc6 26
i.xc6 bxc6 27 :d6 .i.e6 28 ':'xc6 lId8
+, due to Black's bishop-pair.
c) 1 1 'iti>h l lLleS 1 2 h3 ( 1 2 f4? !
lLleg4 1 3 .:tf3 lLlhS ! 14 'ii'e l fS ! was
190 THE GAMBIT GUIDE TO THE MODERN BENONI
slightly better for Black in Zaltsman­
Lobron, New York 1 983) 1 2 . . . gS =
ECO.
d) 1 1 'iVc2 has a number of reason­
able answers such as 1 1 . . .liJb6 and
perhaps even 1 1 . . .liJhS ! ? (Boleslav­
sky's move, made popular by Fischer's
use of it versus Spassky in 1 972), but
the consistent move for us is 1 1 ...liJeS
(D).
Then 12 a4 transposes to the main
line, while White can deviate as fol­
lows:
d l ) 12 h3 gS ! 1 3 liJf3 liJxf3+ 14
i.xf3 liJd7 =. Compare the main lines.
d2) Ambitious, but very loose is
12 f4 ! ? liJeg4 I3 liJf3 ( 1 3 liJc4 liJxe4 !
14 liJxe4 i.d4+ I S c,th l liJxh2! +)
1 3 . . .liJxe4 ! (I think that this is more in­
teresting than 1 3 ...liJhS ! ? 14 h3 liJh6
I S c,th2 fS = Toynt-Shamkovich, USA
1 976, although it's probably no better)
1 4 liJxe4 i.fS I S i.d3 c4 1 6 'iVxc4
nc8 1 7 �b4 as (or 1 7 ...i.xe4 1 8 i.xe4
as =) 1 8 �xd6 .i.xe4 1 9 11i'xd8 ncxd8
20 .i.xe4 :xe4 = 2 1 :d 1 l:te2 22 h3 ! ?
liJf2 2 3 l:td2 lhd2 24 i.xd2 liJd3 ! 2S
i.xaS nxdS 26 i.c3 i.xc3 27 bxc3
liJxf4 +.
d3) Ineffective is 1 2 l:te 1 , when var­
ious moves equalize (for example,
1 2 ...gS; also, since f4 always allows
...�h4, there are lines like 1 2 ...hS 1 3
a4 liJh7), but 1 2 ...liJfg4 ! ? provides a
bit of fun as well: 1 3 h3 liJxf2! 1 4
c,txf2 lIVh4+ I S c,tn i.xh3 ultimately
leads to a draw, as I leave the reader to
work out.
d4) 1 2 b3 liJfg4 ! ? (after 1 2... gS 1 3
i.b2, both 1 3. . .g4 and 1 3. . .liJg6 are
considered equal) 1 3 h3 liJh6 1 4 f4? !
liJeg4! I S liJf3 ( I S i.xg4 i.d4+ 1 6 c,thl
liJxg4 1 7 liJf3 ! liJf6 !) I S ... �a5 16 eS
( 1 6 i.d2 liJe3 !) 16 ...i.fS 17 �d2 dxeS
1 8 hxg4 exf4 1 9 i.b2 liJxg4 intending
...:e3 with a terrific attack, Panczyk­
Kindermann, Polanica Zdroj 1 984.
AI)
11 h3 (D)
B
This is infrequently played, but
very tricky. Now the usual move
1 1 ... liJeS is weak in view of 12 f4. So
Black usually plays . . .
 CLASSICAL MAIN LINE
1 l g5!?
...
The lines that follow this are posi­
tionally ugly for Black, who is de­
pending upon tactics. So we should
definitely consider some calmer ap­
proaches as well. The following alter­
natives deserve attention, and because
they are less forcing and avoid over­
simplification, they may even be the
best way to play for a win:
a) 1 1 .. .'�Jb6 and now:
a l ) 12 a4 i.d7 ( 12 . . . a6 1 3 as tZJbd7
1 4 tZJc4 ! ? tZJxe4 I S tZJxe4 l:1xe4 1 6
tZJxd6 l1d4 1 7 'iVc2 l1xdS 1 8 tZJxc8
.uxc8 1 9 .i.f3 J:d6, probably equal,
strongly resembles line 'c' but without
Black's kingside weaknesses, Pappa­
ceno-Bezviner, Nassau 1 992) 1 3 as
tZJc8 14 lIel ( 14 f4 bS { or 14 . . .:b8 }
IS axb6 tZJxb6 is unclear) 14 . . .:b8 I S
'iVc2 bS 1 6 axb6 axb6 = intending . . .bS
and meeting 17 i.bS? with 17 ...i.xbS
1 8 tZJxbS tZJxdS.
a2) 1 2 l1el a6 1 3 a4 ( 1 3 .i.fl �c7
14 tZJf3 tZJbd7 IS a4 b6 1 6 .i.f4 tZJhS
with the idea 17 .i.h2 .i.b7 1 8 g4 tZJhf6
1 9 eS tZJxeS 20 tZJxeS dxeS) 1 3 ... i.d7
1 4 as tZJc8 I S i.fl tZJa7 1 6 tZJc4 tZJbS
=. We've seen this kind of reorganiza­
tion many times in the Fianchetto Sys­
tem.
It's not clear that White gets any ad­
vantage after 1 1 . ..tZJb6; in any case, it
would be a small one.
b) 1 l . . .h6 (D) is both restrained
and sensible.
Then Black is ready to play . . .gS
without that pawn being exposed in
lines like 1 2 a4 gS or 12 l:1el gS . One
should note that even in lines where
Black plays ... g4, the move . . .h6 is not
191
wasted since it prevents i.gS. Here are
two other continuations:
bI ) 12 �c2 gS 13 tZJc4 ( 1 3 a4 tZJeS
1 4 tZJf3 tZJxf3+ I S i.xf3 is similar to
note 'b' to White's 1 2th move in Line
A2; I S . . . g4 ! ? could follow) 1 3 ... tZJb6
14 tZJxb6 ( 1 4 f4 tZJxc4 I S i.xc4 g4 !
{ I S . . . gxf4 1 6 .i.xf4 tZJhS is unclear;
IS ...tZJhS ! ? } 16 eS !? dxeS 1 7 fxeS l:txeS
1 8 .i.f4 .l:IhS ! +) 1 4 ... axb6 ! ? I S .i.e3
i.d7 1 6 a4 WIIe7 1 7 i.d3 ( 1 7 f3 tZJhS 1 8
�b3 fS) 1 7 . . .g4 1 8 hxg4 tZJxg4 1 9
.i.f4 i.eS 20 �d2 'iHh4 with good
chances.
b2) 12 f4 ! ? is a position from the
9 . . . l1e8 1 0 tZJd2 Four Pawns Attack in
the King's Indian Defence, but with h3
and . . .h6 thrown in. White's h3 seri­
ously weakens g3 in lines where Black
plays ...tZJe4 or ... tZJhS, while the move
... h6 prevents i.gS, a common idea for
White. On the other hand, h3 prevents
the move ... tZJg4, a tactical leap that
Black uses in many lines. The reader is
invited to compare the normal tZJd2
Four Pawns variations (which are
briefly discussed in note 'b' to White's
1 1 th move in Line A).
192 THE GAMBIT GUIDE TO THE MODERN BENONI
b2 1 ) I believe that Black probably
does about as well as usual in the lines
with 12 ...c4; here are just a few brief
ideas to give an idea of the play, with
both sides having alternatives on more
or less every move:
b21 1 ) 1 3 a4 �c5 14 e5 dxe5 1 5
�xc4 exf4 ! 1 6 .i.xf4 �ce4 =.
b2 1 2) 13 �hl �c5 14 e5 dxe5 1 5
fxe5 1:.xe5 1 6 �c4 1:.e8 (or 1 6 . . .1:.f5
1 7 i.f4 g5 ! 1 8 i.e5 �fe4 =) 1 7 i.f4
(White would normally play 17 i.g5
here) 1 7 . . . �ce4 1 8 i.f3 iH5 with
equality.
b2 1 3) 1 3 i.xc4 �c5 14 �c2 ( 1 4
1:.el �h5 =; 14 e5 dxe5 1 5 fxe5 1:.xe5
1 6 �f3 Ue8 =) 14 ...�h5 1 5 i.b5 .i.d7
1 6 .i.xd7 �xd7 17 �h2 �e7 1 8 1:.f3
1:.ac8 =.
b22) Actually, I think that Black
should probably toss in 1 2 ... a6 1 3 a4:
b22 1 ) If Black continues 1 3 . . .c4,
he has prevented �b5 or i.b5 and
opened up more ideas, especially after
1 4 i.xc4 - compare lines under 'b2 1 ' .
b222) Black can directly make use
of ... h6 by 1 3 ...�h7 ! ? The idea is to
play for ...g5 and/or ...�df6. For ex­
ample, 14 �c4 ( 14 �h l can be met by
1 4...�df8 1 5 .i.f3 g5 ! ? or 1 4... i.d4 ! ?
1 5 �b3 ! ? i.xc3 1 6 bxc3 ':xe4 1 7
i.d3 1:.e8 intending 1 8 f5 g5) 1 4. . .�e7
(or just 1 4 ...�b6 =) 1 5 i.f3 g5 1 6 i.e3
gxf4 1 7 .i.xf4 �e5 1 8 �b6 ':b8 1 9
�d2 �g5 20 .i.h5 �g6 ! '
b23) 1 2 . . .1:.b8 1 3 <it>h l �b6 ! ? i s
noteworthy; e.g., 1 4 a4 a6 1 5 a5 �a8
intending ... �c7.
I don't think that Black need fear
1 1 ... h6 1 2 f4, but it certainly makes for
interesting play.
c) l l .. .a6 1 2 a4 ( 1 2 �c4 �xe4 1 3
�xe4 1:.xe4 1 4 �xd6 Ud4 + ; 1 2 f4 b5)
12 ... g5 ! ? 13 �c4 �xe4 14 �xe4 ':xe4
1 5 �xd6 1:.d4 1 6 'ii'c2 (D).
B
1 6 ...':xd5 ! ? (compare the main line
{ i.e. without ... a6 and a4 } , in which
Black plays ... �b6 in an equivalent
position) 17 �xc8 Uxc8 1 8 i.f3 ltd6
1 9 i.xb7 1:.b8 20 i.e4 h6. Black is
structurally bankrupt, but he has pres­
sure down the b-file, more active
pieces, and, soon, a dominating bishop
on d4. That's probably still not enough
to secure complete equality against
White's powerful bishops, but here are
some sample lines:
c l ) 21 f4 i.d4+ 22 �h2 g4! ? 23
1:.a3 ! h5 (23 ...�h4) 24 .i.f5 �h4 25
Ug3 �f6 26 b3 1:.e8.
c2) 21 1:.bl i.d4 and now:
c2 l ) 22 h4?! �f6 23 i.f5 (23 hxg5
�xe4 24 �xe4 hxg5 =) 23 ...�d5 24
hxg5 hxg5 25 i.d2 1:.h6 26 g3 (26
�c 1 'ii'd6) 26...�f6 27 b3 �g7 ! 28
'ii'e4 �e7 29 .i.d7 1:.bh8 -+.
c22) 22 �h l �f6 23 i.f5 ltdb6 24
f4 (24 b3 �d5 +) 24...�h5 ! +.
 CLASSICAL MAIN LINE
c23) 22 .i.e3 .i.xe3 23 fxe3 liJf6.
c24) 22 .i.f5 liJe5 23 b3 liJc6 :t.
c25) 22 b3 liJf6 23 i..f5 .:!.db6 24
h4 1i'd5 25 hxg5 hxg5 26 i..xg5 l1xb3
27 i..xf6 i.. xf6 28 l1xb3 l1xb3 :t.
In general, Black should not be
optimistic about achieving full theo­
retical equality in this line, but his po­
sition is better and more resilient than
one might have supposed. The main
line with 1 1 ...g5 is probably more reli­
able.
12 liJc4!
For 1 2 a4 liJe5, see note 'b' to
White 1 2th move in Line A2. After 1 2
liJf3, 1 2 ... g4! is fine.
The best alternative is 12 ':'e1 liJe5
1 3 liJf1 h6 1 4 liJg3 a6 1 5 a4, when I
like 15 ... g4 ! ; e.g., 16 hxg4 liJfxg4 1 7
i..f4 ( 1 7 f3? 1i'h4) 1 7 . . .1i'h4 1 8 1i'd2
llVh2+ 1 9 <ittf 1 h5 20 i..f3 i..d7 with an
obscure position that looks promising
for Black.
12 liJxe4 13 liJxe4 .:!.xe4 (D)
•••
w an awful sight and White has two good
14 liJxd6
14 .i.d3 11h4! 1 5 liJxd6 liJe5 1 6 liJf5
(this position first arose in the game
193
Donner-Hartoch, Wijk aan Zee 1 972)
16 . . .liJxd3 ! 17 1i'xd3 and instead of
theory' s 17 ... c4, I think that 17 ....i.xf5
1 8 1i'xf5 ':'d4 1 9 .i.xg5 llVxd5 is best.
14 .:!.d4 15 �c2 liJb6!?
.••
One of two methods here. The other
is 1 5 ....:!.xd5 1 6 liJxc8 .:!.xc8, but since
including . . . a6 and a4 probably im­
proves Black's chances, I refer you to
note 'c' to Black's 1 1th move.
16 liJxcS l1xcS 17 i..e3 ':'b4
The exchange sacrifice 17 ...liJxdS 1 8
.i.xd4 i..xd4 was advocated b y Nor­
wood in his book, but 1 9 1i'f5 ! is es­
sentially a refutation.
IS .i. g4 ':!'xb2
The immediate 1 8 . . ..:!.c7 1 9 ':!'ad 1 !
ltxb2 20 �f5 ! .i.f6 2 1 d6 is worse for
Black.
19 �cl
Here my ever-alert editor pointed
out that 1 9 ... ':'c7? 20 i.. xg5 �xd5, as
in Lahav-Psakhis, Tel-Aviv 1990, al­
lows 2 1 ltel ! , which practically win­
ning. Thus, Black's best chance is ...
19 l1bS!
.••
... when his own weaknesses make
bishops to boot. But Black is counting
upon his extra pawn, piece activity
and threats against the d-pawn to con­
fuse the issue. White can hardly be un­
happy with his structure, but still
needs a way to neutralize Black's ini­
tiative. Here are some lines:
a) 20 i..xg5 llVxd5 21 ltd l (2 1 i.. f3
llVf5 ; 2 1 llVf4 i..e5 22 llVf5 .i.d4)
2 l .. ..i.d4.
b) 20 i.. xc5 .:!.b5 2 1 .:!.b1 ltxbl 22
llVxb1 llVxd5 is about equal.
c) 20 .:!.d l ! ? .:!.b4 ! ? 2 1 i.. xg5 1i'd6.
194 THE GAMBIT GUIDE TO THE MODERN BENONI
d) 20 'iWxcs 'iWxdS 2 1 'iVc7 (21 'iVe7
h6 22 :ad1 'iWeS 23 :d8+ ':xd8 24
'iWxd8+ �h7; 2 1 'iVxdS liJxdS 22 i..xgS
h6 23 i..h4 :e8 { 23 ...liJc3 ! ? } 24 �He 1
lhe1 + 2S 1he1 i..f6 should be drawn)
2 l ...i..eS (or 2 l .. ..l:.e8 ! ?, and then 22
i..f3 'iWe6 23 i..xb7 h6 24 llfe1 liJc4 or
22 ':fd 1 'ilic6 23 ':ac 1 'iWxc7 24 .l:.xc7
i..eS 2S ':xb7 liJc4 26 ':xb2 liJxe3 = )
22 'iWe7 h6 23 i.hS (23 ':fd1 'iWe4)
23 . . . 11bS 24 .:tfd1 'iWe6 2S 'ilixe6 fxe6;
the combination of Black's extra pawn
and White's bishops will very likely
lead to a draw.
This has been a rather long look at
the underanalysed 1 1 h3. Black has
the choice between the active but
anti-positional variations 1 1 . . . gS or
1 l . ..a6 12 a4 gS, of which the former
seems better, and the solid ideas of
1 l . ..h6 and 1 l .. .liJb6, both of which
need tests.
A2)
11 a4
The normal choice. White wants to
clamp down on Black's queenside ex­
pansion without delay, and as a side
benefit, he gets the option of a rook­
lift to a3 at some point.
11 ...liJe5 (D)
This is the key position for the
. . .liJbd7 defence, which arises in most
of the games. Black will normally play
... gS next, to secure the knight on eS
against f4. White's most important
idea, liJc4, may now be answered by
...liJxc4, after which Black may reoc­
cupy eS by ... liJd7-eS or ... liJg4-eS.
In general, Black pursues a dark­
square strategy on the kingside and in
w
the centre. We shall see, for example,
. .. liJg6-f4 or, after the natural move 0,
. . .liJhS-f4. Sometimes moves such as
. . .i.eS (or . . . i.d4) and ...'ilVf6 will rein­
force that theme. If he deems it neces­
sary to counteract ideas like f4 or even
liJo, Black may continue with ... g4 in
order to restrict White's kingside. In
that case . . . liJhS-f4 can be played
without a preliminary f3 on White's
part, and Black's queen can join the
attack from h4 or gS.
The strategy just outlined is un­
usual in the Benoni, and this is cer­
tainly the first time that we have seen
such a piece arrangement as Black's
primary defensive method. Its main
drawback is the surrender of the fS­
square to White; a knight posted on fS,
for example, can both disrupt Black's
game and serve to block his aggressive
plans. A related problem is that White
might enforce f4 despite Black's ef­
forts to prevent it by . . .gS . After . . . gxf4
Black gains a permanent outpost on
eS, to be sure, but White's f-file (and
further control of fS) can also be influ­
ential. I think that what follows will
 CLASSICAL MAIN LINE
demonstrate how these ideas can be
successfully counteracted.
The . . . ttJe5/. . . g5 plan is more fun
and requires more skill than the ...ttJa6-
c7 schemes in Line B, but it is also
riskier. The reader may want to com­
pare these two systems before decid­
ing upon a personal repertoire.
12 'ii'c2
This remains the most popular
move for White. 1 2 'iVc2 reinforces e4
and thus allows the move ttJc4. It also
prepares b3 and i.b2 by protecting the
knight on c3, and opens up d l for a
knight, since the idea of ttJe3-f5 is log­
ical after . . . g5 .
However, in this case the alterna­
tives are roughly as good, and must be
looked at very carefully:
a) 1 2 ttJdb l ! ? h5 ! ? ( 1 2 . . . ttJfg4 ! ?
1 3 ttJa3 f5 = ; 1 2 .. :�e7 Nunn) 1 3 h3
ttJh7 ! (an idea we have seen before,
this time tactically based) 14 f4 ttJg4 !
15 'fUel , Traito-Agapov, USSR 1 987,
15 ... i.d4+ ! 16 �h l ttJgf6 17 i.f3 ( 17
ttJd2 i.xc3 +) 17 . . . i.f5 (+ Agapov;
Schneider disagrees, but I think White
really is in trouble) 1 8 �e2 ttJxe4
( 1 8 ... i.xc3 1 9 ttJxc3 ttJxe4 20 ttJxe4
i.xe4 2 1 i.xe4 ttJf6 is very strong for
Black) 1 9 ttJxe4 i.xe4 20 i.xe4 and
now Schneider's line continued 20. . .f5
2 1 ttJc3 ttJf6 22 'iVb5 fxe4 23 f5, but
Black should prefer 20 ...ttJf6 +.
b) 1 2 h3 g5 1 3 ttJf3 ttJxf3+ ( 1 3 . . . g4
14 ttJxe5 ':'xe5 has also been played)
14 i.xf3 h6 1 5 hlel ( 1 5 %:tbl b6 1 6 b4
ttJd7 1 7 ttJb5 ttJe5 1 8 i.h5 a6 1 9 ttJc3
ttJg6 20 'iVc2 %:ta7 ! = Schneider; 1 5
lia3 b6 1 6 �c2 a6 1 7 ttJe2 ! i.d7 1 8
ttJg3 b5 = Tisdall-Amason, Brighton
195
1 98 1 ) 1 5 . . . a6 (here this move is not
bad, because enforcement of . . . b5 or
control of the b-file is the next consis­
tent step; 1 5 . . .b6 ! ? 1 6 �c2 a6 is the
solid alternative) 16 a5 l:tb8 17 i.d2
i.d7 = Tunik-Pigusov, Russian Ch (St
Petersburg) 1 998.
c) 12 :a3 goes well with a4, threat­
ening to swing over to the kingside in
many lines. It is useful, however, that
Black has not spent a tempo on ... a6,
and thus has an extra tempo with which
to implement his dark-square strategy:
12 ... g5 13 :e l ( 1 3 �c2 transposes to
the main line; 1 3 ttJf3 ttJxf3+ 1 4 i.xf3
g4 1 5 i.e2 ttJxe4 1 6 i.xg4 is com-
fortable for Black after 1 6 . . . �h4 or
1 6 . . . ttJxc3 and . . .�h4) 1 3 . . . ttJg6 1 4
i.b5 .l:.e7 (also possible i s 14 . . .:f8;
note that Black avoids ... i.d7 and the
exchange of bishops because that
would give up protection of f5 - yet if
a white knight is far from f5, he might
allow this for the sake of taking over
light squares following ... c4 and ... ttJc5
or ...ttJe5) 1 5 ttJn a6 1 6 i.d3 ( 1 6 i.c4
h6 1 7 tbg3 i.d7 1 8 i.d2 �c7 1 9 'iNc2
l:lae8 20 a5 ttJg4! 2 1 h3 tb4e5 22 i.n
ttJh4! '+ intending .. .f5, Stempin-Suba,
Prague 1 985) 16 . . .h6 1 7 ttJg3 :b8 ! ?
( 1 7 .. :�c7) 1 8 a5 �c7 1 9 i. n i.d7 20
i.d2 l:.be8 ! = A.Petrosian-Anikaev,
Telavi 1 982.
d) 12 :e l reinforces e4 and pre­
pares ttJn , but it takes away a de­
fender of f2 and renders an early f4
less likely. Black plays 1 2 . . . g5 (D),
when White has tried these moves:
d 1 ) 1 3 ttJc4 ttJxc4 14 i.xc4 ttJg4
1 5 h3 ttJe5 1 6 i.b5 l:H8 1 7 i.e2 �h8 =
Kraidman-Tisdall, Gausdal 1 983.
196 THE GAMBIT GUIDE TO THE MODERN BENONI
d2) 1 3 .tb5 '!J.e7 (exchanging light­
squared bishops would leave f5 very
weak, especially with lDn -g3-f5 in
the air) 14 lDn h6 1 5 lDg3 lDg6 1 6
.lid2 lDg4 1 7 h3 lD4e5 1 8 lDh5 .lih8
19 f4 gxf4 20 lDxf4 a6 21 .lin �h7 22
lDh5 (Schneider gives 22 'ii'h5 ' ! ' .lig7
23 a5, continuing 23 .. J:te8 24 .ie2
'ii'h4 25 'iYxh4 ';!;' , but either 23 . . .b5
24 axb6 'iYxb6 or even, at the end of
Schneider's line, 25 ... b5 looks fine for
Black) 22 . . .Ite8 23 g3 (23 'ii'c l �g8 24
.ixh6 lDf3+! 25 gxf3 lDe5+ 26 �h l
lDxf3 -+) 23 ...lDh4 ! 24 �e3 (24 gxh4?
l:tg8+ 25 .ig2 .lixh3 26 lDf4 'ii'xh4
gives Black a winning attack) 24. . .'ii'g5
(or 24 .. .l:1g8) 25 �h l lDhf3 26 �xf3
'iYxh5 27 lH4 'iYg6 28 l:Ih4 (28 'iYel
.:tg8 29 g4 lDxg4 ! 30 hxg4 .txg4 3 1
'iYf2 i s given by Schneider, but then
Black has 3 l . ...lie5 ! +) 28 . . .�xg3 29
.l:i.xh6+ �g8 30 'ii'e l 'ii'xe l 3 1 .ixel
lDg6 32 .id2 l::te5 =+= Ftacnik-Wom­
acka, Bundesliga 1990/1 .
d3) 1 3 lDn (the most direct) 1 3 ...h6
14 lDg3 ( 1 4 a5 a6! 1 5 lDe3 l1b8 =)
14 ... lDg6 15 a5 ( 1 5 .id2 lDf4 16 .ib5
Ite7 1 7 �c2 lDg4 = Stohl-Kindermann,
Stary Smokovec 1986) 1 5 ...a6! 1 6 .id2
I:Ib8 1 7 'ii'c2 lDf4 1 8 .ic4 lDg4 1 9
lDce2 'ii'f6 = Gyimesi-Pigusov, Kosz­
alin 1 999, having in mind 20 lDxf4
gxf4 2 1 .ic3 fxg3 ! 22 .ixf6 gxf2+ 23
�xf2 lDxf2 24 .ixg7 lDh3+ with a
slight advantage for Black.
12 ... g5 (D)
13 .l:i.a3
Considered best by Schneider, and
the main line in ECO. The move l:ta3
appears in several variations of the
Benoni, but never so prominently as in
the Classical lines. It serves primarily
for defence of the kingside. Described
in the style of Nimzowitsch, the com­
mander of White's army pokes his
head over the hill to assess the strength
of Black's attack on his monarch. He
realizes that his own forces are them­
selves ample for defence if he rushes
them across the plains in time. His di­
visions will not only serve in defence,
but in the case of a reckless enemy ad­
vance they will be used to launch a
counterattack against the now-exposed
king. It is therefore up to Black's forces
 CLASSICAL MAIN LINE
to improve their own positions rather
than rush precipitously forward.
Back on the chessboard, several of
White's options echo the themes in the
previous note. Of these, 1 3 ttJd l and
1 3 ttJc4 are the most important:
a) 1 3 b3 g4 (or 1 3 ... b6 14 iLb2 a6
IS ttJdl J:la7 ! 1 6 ttJe3 l:tae7 = Vrane­
sic-Tarjan, Chicago 1 973) 14 iLb2 ttJhS
( l 4. . . a6 I S as iLd7) I S g3 ( 1 S ttJc4
ttJxc4 1 6 iLxc4 ttJf4 17 ttJd 1 iLeS =
Gligoric-Tatai, Venice 1 97 1 ), and here
I like simply I S ... ttJf6 ! =, now that g3
has weakened White's kingside.
b) 1 3 ttJf3 ttJxf3+ 14 iLxf3 ttJg4
(or 14 ...ttJd7 I S i.g4 ttJeS 16 iLxc8
l:xc8 = intending . . . c4 and . . .ttJd3, as
given by Nunn) I S iLxg4 iLxg4 1 6 f4
gxf4 1 7 iLxf4 iLhS 1 8 ttJbS iLeS =
Kharitonov-Chekhov, Yaroslavl 1 982;
Black can pressure the e-pawn.
c) 1 3 ttJd l is unmentioned in ECO,
but nevertheless merits attention:
c 1 ) 1 3 . . .g4 14 ttJe3 ( 1 4 iLbS ':'e7
IS l:le l { I S ttJe3 ttJg6 ! ? } I S ... a6 1 6
..tfl b6 ! ? =) 14 . . . a6! ( 1 4. . .ttJhS ? ! I S
f4 ! gxf3 1 6 ttJxf3 ttJf4 17 ttJxeS iLxeS
1 8 ttJg4 with advantage - Psakhis) I S
f4 ! ? ( 1 S J:l e l ttJg6 =) I S . . . gxf3 1 6
ttJxf3 ttJg6 17 ttJfS ? ! iLxfS ! 1 8 exfS
l:t:,e7 1 9 iLgS 'iVd7 and Black is
slightly better.
c2) 1 3 . . . ttJg6 keeps pressure on e4:
14 ttJe3 ( 1 4 f3? ttJxdS ; 14 iLbS l:te7 I S
ttJe3 a6 1 6 iLc4 iLd7 { or 16. . .g4 } 1 7
ttJfS J.xfS 1 8 exfS ttJeS 1 9 iLe2 g4 =)
14 ... a6 ( 1 4. . . ttJf4 I S iLbS l:te7 16 J:lel
g4 is unclear) I S l:tel g4 1 6 J:lbl (Pel­
aez-R.Grinburg, Buenos Aires OL
1 978) 16 ... ttJf4 ! 1 7 b4 ( 17 iLfl 'iVe7 1 8
b4 ttJxe4 +) 17. . .ttJxe2+ 1 8 ':xe2 b6 1 9
197
bxcS bxcS 20 iLb2 1i1 a7 ! = intending
. . .J:lae7.
d) After 1 3 ttJc4 ttJxc4 14 iLxc4
Black seems to have a number of satis­
factory options:
d l ) 14 ... ttJd7 IS ttJe2 a6 16 ttJg3 ( 1 6
a s ttJeS 17 iLb3 iLd7 ! =) 16 . . . ttJeS 17
iLe2 :b8 1 8 f4 gxf4 1 9 iLxf4 ttJg6 20
iLd2 iLeS 2 1 iLc3, Kratochvil-Poloch,
Ceske Budejovice 1 999, and here
2 1 . . .iLxg3 ! 22 hxg3 'iVgS gives Black
at least equality in view of White's
weaknesses.
d2) Also playable is 14 ...ttJg4 I S
ttJe2 'iWe7 1 6 ttJg3 iLd4 ! 17 h3 'ilVeS 1 8
.l:.a3 ttJf6 = 1 9 :f3 ! ? g4 20 hxg4 ttJxg4
2 1 iLf4 'iWf6 22 ttJe2 iLeS 23 iLxeS
'iYxeS = O.Foisor-Suba, Romania 1 984.
d3) 14 . . .ttJhS I S ttJe2 ( 1 S g3 iLh3
16 1i1e 1 iLeS 17 'iVd 1 ttJg7 1 8 iLbS
,iH8 =, preparing ... fS, Flear-Plaskett,
Lewisham 1 983) I S . . .'i!Vf6 ! ? (a rare
but appealing move) 16 l:ta3 'i!Vg6 1 7
ttJg3 ttJf4 1 8 f3 iLeS 1 9 'it> h 1 'iVh6 and
Black is quite OK, Barbero-Wesseln,
Delmenhorst 1 986.
We now return to the position after
1 3 1i1a3 (D):
B
198 THE GAMBIT GUIDE TO THE MODERN BENONI
13 g4
...
This is ambitious, but by no means
the only concept that Black can pur­
sue. In many such positions, for exam­
ple, Black plays more conservatively
by ... h6, when f4 at some point results
in the trade-off of Black's e5 outpost
versus White's f-file. The following
tries don't as yet commit Black to a
specific structure:
a) 1 3 . . .�h8 ! ? 14 tbc4 ( 1 4 h3 g4)
14 ... tbfg4 1 5 tbxe5 tbxe5 16 f4 gxf4
17 .ltxf4 a6 ( 1 7 . . ..l::.g 8 looks more use­
ful) 1 8 'iYd2 .ltf6 1 9 'it>h 1 lIb8 20 h3
(20 tbd I l:tg8 hopes for 2 1 .ltxe5
.ltxe5 22 l:txf7 l:txg2 ! , although here
2 1 tbe3 ! would keep the advantage)
20 . . .tbg6 2 1 .lth6?! (Timoshenko pre­
fers 2 1 .lth2 "fie7 22 tbd l , when he
calls 22 . . . .lte5 23 l:taf3 "unclear";
maybe 22 ... .ltd4, to keep . . .tbe5 as an
option?) 21 . . .l:tg8 22 tbdl tbe5 23 tbf2
.ltd7 24 .ltf4 "fie7 + Timoshenko­
Lobron, Moscow 1 989.
b) 13 . . .1i'e7 is a safe option, proba­
bly better than 1 3 . . . �h8: 1 4 a5 h6
( 1 4. . . g4 1 5 f4 gxf3 1 6 tbxf3 tbxf3+ 1 7
..txf3 tbg4 1 8 .ltxg4 ..txg4 1 9 .ltf4 is
"unclear" according to Gelfand; Black
looks fine) 1 5 tbd l ? ! ( 1 5 .ltb5 11f8 1 6
tbd l Gelfand) 1 5 . . . .ltd7 = Gelfand-
1. Hall, Malmo 1 999; Black plans
. . .tbg6 next.
14 tbc4
Here we have the usual suspects:
a) 14 l:te l tbh5 1 5 tbn 'iYf6 1 6
tbd l 'iYg6 (this idea again, to enforce
. . .f5 and keep an eye on the e-pawn)
17 .ltb5 :f8 1 8 tbde3 a6 19 .lte2
tbf3+! 20 gxf3 gxf3+ 2 1 tbg3 tbxg3
22 hxg3 fxe2 23 tbf5 .ltxf5 24 exf5
'iYf6 (24 . . .1Wg4 25 �xe2 .ltd4 26 �g2
nfe8) 25 :b3 'iWe5 26 l:txb7 ':'fe8 27
nb6 'iNxd5 28 lIxe2 'iVf3 29 l:te3 1he3
30 .ltxe3, Zaltsman-Y.Grunfeld, Lone
Pine 1 98 1 , and now 30. . ..lte5 is at least
equal.
b) 14 b3 tbh5 ( l 4 . . . a6 may also
suffice: 15 a5 tbh5 16 tbc4 "iWf6 1 7
tbd I ? { I 7 tbb6 :b8 1 8 tbxc8 .l:.bxc8
= } 17 . . . tbf3+ 1 8 gxf3 gxf3 1 9 .ltxf3
'iYxf3 20 b4 'ii'xe4 + Vefling-Charu­
shin, corr. 1985) 1 5 tbc4 �h4 ! (D) and
now:
b I ) Not 1 6 tbxd6? tbf3+ 17 .ltxf3
( 1 7 gxf3 .lte5 1 8 f4 tbxf4!) 17 . . . gxf3
when 1 8 tbxe8?? fails to 1 8 . . .'iWg4.
b2) 16 tbb5 tbxc4 17 bxc4 �e7 1 8
.ltd3 (here Schneider claims a posi­
tional advantage for White, but I doubt
it, since the situation is very similar to
what we've seen above) 1 8 . . ...te5 1 9
.ltb2 ( 1 9 f4? gxf3 20 �hf3 a6 2 1 tbc3
.ltg4) 1 9 . . . .ltxb2! (instead of the com-
plex 1 9 . . . f6 ! ? of A.Schneider-Bonsch,
Budapest 1 982) 20 'iVxb2 tbf4 2 1
'iVd2 ! tbg6 = 22 a5 (22 'iVh6 a6 23
tbc3 'ii'f8 =) 22 . . .�g7 23 tbc3 tbe5 24
 CLASSICAL MAIN LINE
�b l fS ! , having in mind 2S exfS ?
lDxd3 26 f6+ (not 26 'ii'xd3? i..xfS !)
26. . :iVxf6 27 'ii'xd3 bS ! =t.
c) 1 4 lDd l lDhS ! ? (safer is 14...tDg6
I S lDe3 'ike7 1 6 i..bS 1:I.f8 1 7 as l:tb8
1 8 ':'dl h6 1 9 l:tb3 lDh7, Averkin-Pigu­
sov, Kazan 1 985) I S f4 ( 1 S lDe3 lDf4
1 6 i.. d 1 b6 = Antunac-Y.Griinfeld,
New York 1 979) IS . . .gxf3 16 ttJxf3
lDg6 1 7 lDgS lDf6 and now, instead of
1 8 .:tg3, as in Kouatly-Renet, Mar­
seilles 1 988, Schneider likes 1 8 .:taf3
h6 1 9 1:.3f2, but I see no major prob­
lem after 1 9 . . .lDeS; e.g., 20 b3 a6 2 1
lDh3 .ltxh3 22 gxh3 lDh7 with the idea
. . .lDgs.
14 lDh5
•••
After 14 . . .lDxc4 IS i..xc4, IS ...ltJd7 !
looks more natural than I S . . .lDhS 1 6
g 3 ;1;, a s has been tried i n practice.
Then after . . . lDeS, the ideas of . . .'ikM
and .. .fS are still available later.
15 lDxe5 i..xe5 16 g3 lDg7
Better may be 1 6 . . . i..d 7, and if
White proceeds as in the game with 1 7
lDd 1 , then Black can play (among oth­
ers) 17 . . .'ii'f6 1 8 f4 ( 1 8 lDe3 'ii'g6 1 9
lDfS .ltxfS 20 exfS 'ikg7 and ...ttJf6)
1 8 . . . gxf3 1 9 .l:.axf3 'ii'g6 with coun­
terplay.
After the text-move ( 1 6...lDg7), Vil­
ela-Vera, Havana 1 987 continued 1 7
i..d3 a6 ! ? ( 1 7 . . ..ltd7 ! ?) 1 8 ttJ d I ! i.. d7
1 9 as i..bS ( 1 9 . . .bS 20 axb6 'iixb6 2 1
lDe3 ;1;) 20 ttJe3 'iid7 2 1 lDc4 i.. xc4 22
'ii'xc4. This is probably somewhat
better for White, but this whole line is
not worrying in view of the multiple
alternatives for Black along the way.
This concludes our look at 9 . . .l:I.e8
1 0 lDd2 lDbd7 in the Classical Main
199
Line. I think it offers Black level play
in an unbalanced position.
B)
7 lDd2 i..g7 8 e4 0-0 9 .lte2 l:te8 10
0-0 lDa6 (D)
This introduces a totally different
scheme than we saw with 1O . .ttJbd7
.
(Line A). Black's initial idea is very
simple. He will play . . .lDc7, discour­
aging an eS break by White because
the pawn on dS will hang. The knight
on c7 also helps to support . . . bS, usu­
ally prefaced by . . .a6. If White plays
a4-aS, this knight frequently exercises
the option of moving to bS and from
there to d4.
You will find this to be an unpreten­
tious system, which offers Black a
clear plan and reliable play without
creating the weaknesses of Line A
above. On the other hand, he tends to
get fewer winning chances, since both
sides' play is generally less dynamic.
1 1 f3
The most common move. Unlike the
. . . lDbd7 system of the last section, the
200 THE GAMBIT GUIDE TO THE MODERN BENONI
alternatives to the popular main line
aren't as serious (i.e., equal in value to
it):
a) 1 1 i.. xa6 is hardly ever effective
in the Benoni, since Black gets two
bishops and strong b-file pressure; for
example, l l ...bxa6 1 2 .:tel i..d7 (or
1 2 . . .lLlg4 1 3 h3 lLleS 14 'iVe2? fS ! I S
b 3 f4 ! 1 6 i..b2 g S with an attack,
Kholmov-Sandarov, Riga 1 9S4) 1 3 a4
lLlg4 14 h3 lLleS I S lLlf1 .:tb8 with an
equal position, Kaestner-Neumann,
Wiesbaden 1 992.
b) 1 1 f4 transposes to a slightly un­
usual variation of the Four Pawns At­
tack, the main line going 1 l . ..lLlc7 1 2
a4 ( 1 2 i..f3 l:tb8 1 3 lLlc4 b S 14 lLlxd6
'ii'xd6 I S eS 'iVb6 1 6 exf6 i..xf6 +')
1 2 . . .b6 ( 1 2 ... a6 is also supposed to
lead to equality, according to ECO) 1 3
'iith 1 ( 1 3 l:tel :tb8 1 4 i..f3 i..a6 = )
1 3 . . ..i.a6 14 :tel ( 14 i.. xa6 lLlxa6 in­
tending . . . lLlM) 14 . . . i.. xe2 I S 1::txe2
lLlg4 16 h3, Savon-Beliavsky, USSR
Ch (Moscow) 1973, and now 16 ... i..d4!
looks best; e.g., 1 7 'ii'e 1 lLlf6 1 8 lLlc4
( 1 8 lLlf3 'ii'e7 19 eS i.. x c3) 1 8 ...'ii'e7
( 1 8 ... 'ii'd7 =) 19 eS? dxeS 20 d6 'ii'e6
2 1 fxeS 'ii'xc4 22 dxc7 .i.xc3 23 bxc3
lLldS =t.
c) 1 1 a4 lLlM 12 .:tel ( 1 2 f3 lLld7
1 3 lLlc4 lLleS =) 12 . . .b6 1 3 .:ta3 'ii'e7 14
lLlc4 i.. a6 I S i..f1 i.. xc4 16 .i.xc4 a6 =
Temirbaev-Ruban, Uzhgorod 1 988.
d) 1 1 'iii'h i lLlc7 1 2 a4 :tb8 13 f3
a6 ! ? ( 1 3 ... b6 transposes to note 'a2' to
White's 1 3th move) 1 4 lLlc4 bS I S
axbS axbS 16 lLlaS i..d7 1 7 lLlc6 .i.xc6
1 8 dxc6 b4 1 9 lLldS? ( 1 9 lLla4 lLle6 ! =)
19 . . .lLlfxdS 20 exdS lLlxdS 21 'ii'xdS
1:txe2 and Black had the better of the
complications in Shneider-Agzamov,
Telavi 1 982.
e) 1 1 :tel (a fairly frequent move,
but ineffective at this point for several
reasons) 1 l ...l2Jc7 ( 1 l .. .:tb8 has also
been successful here) 12 a4 a6 (less
demanding than the popular and also
satisfactory 12 . . . b6) 1 3 'ii'c2 ( 1 3
:tbl ! ? l:tb8 14 M cxb4 I S :txM lLld7
16 'ii'c2 lLlcs =) 13 ...l:tb8 14 as .i.d7 I S
lLlc4 lLlbS 1 6 i..gS lLld4 17 'ii'd 1 i..bS
is level, Makogonov-Antoshin, Baku
1 9S9.
1l lLlc7 12 84 (D)
•••
After 1 2 'ii'c2 Boleslavsky recom­
mends 12 . . .bS, continuing 1 3 lLlxbS
lLlfxdS =, but 1 3 .i.xbS ! lLlxbS 14
lLlxbS is good for White in view of
14 ...i.. a6 I S 'ii'a4 'iVb6 1 6 lLlc7 ! . So
Black should play something sensible
like 12 . . . l:tb8, or try 1 2 ... lLlfxdS ! ? 1 3
exdS .i.xc3 14 'iWxc3 ':xe2 I S lLle4
lLlbS ! 1 6 'iVd3 lLld4 17 i..e3 l:txb2 1 8
i..xd4 cxd4 1 9 'ilVxd4 'iVb6, which is
probably about equal.
B
12 b6
•.•
Or:
 CLASSICAL MAIN LINE
a) 1 2 ...lLld7 ! ? has done well over
the board and the reader may prefer it,
but I don't fully trust the line 1 3 lLlc4
(for 1 3 'iii>h l b6, see note 'ai ' to
White's 1 3th move; 1 3 f4 is consid­
ered equal after 13 ... lLlf6, but simpler
may be 1 3 ... fS ! ? 14 exfS gxfS I S lLlf3
ctJf6 1 6 �d3 lLlg4 1 7 h3 lLle3 1 8 �xe3
':xe3 = Baburin-de Firmian, Farum
1 993) 1 3 ...lLleS 1 4 lLle3 fS I S exfS !
( 1 S f4? ! lLlf7 1 6 exfS lLlh6 ! ! 1 7 fxg6?
i.d4 1 8 gxh7+ 'iii>h8 1 9 l:tf3 ctJg4 20
'iVd3 'it'h4 2 1 h3 ctJxe3, Toth-Matulo­
vic, Hungary 1 972, and now the crit­
ical continuation 22 �xe3 is refuted
by 22. . . lIxe3 23 lIxe3 'it'xf4 24 lLld l
lLlxdS) I S ... gxfS 16 f4 lLlf7 ( 16 ...ctJg6
has also been played, but 1 7 'iii>h l pre­
pares to pressure the f-pawn) 1 7 'iii>h I !
( 1 7 �d3 �d4 ! ? 1 8 :f3 'it'f6 1 9 'iii>h l
.i.d7 20 �d2 lLlh6 2 1 'it'c2 l:te7 = is
ECO's main line) 17 ... 'it'f6 18 �d3
i.d7 1 9 'it'c2 ctJh6 20 �d2 lIf8 2 1
l:tf3 ! , Hesselbach-Mueller, COIT. 1 98 1 .
White must be at least somewhat
better due to the weak fS-pawn.
b) It's curious that 1 2... a6 ! ? is never
even mentioned, presumably because
1 3 ctJc4 lIb8 14 �f4 �f8 looks so bad.
But it seems to me that White has to
play quite accurately to maintain an
edge, probably by I S as ( 1 S 'itd2 bS
1 6 ctJaS �d7 is not clear) I S ... ctJbS
( l S ...ctJhS ! ? 1 6 �e3 �d7 merits con­
sideration) 16 ctJxbS axbS 1 7 lLlb6 ( 1 7
ctJa3 ! ?) 1 7. . .�d7 ( 1 7 . . .ctJhS 1 8 �e3
i.d7) 1 8 'it'd2, intending to answer
1 8 ...ctJhS with 1 9 �gS, and thinking
about a timely b4.
13 lLlc4
Alternatives look too slow; e.g.:
201
a) 13 'iii>h l and now:
a l ) 1 3 . . .lLld7 ! ? 1 4 lLlc4 lLle5 1 5
lLle3 :l.b8 1 6 �d2 a6 1 7 f4 lLld7 1 8
lLlc4 lLlf6 1 9 �f3 b5 20 lLlxd6 ! 'it'xd6
2 1 eS, Dijksta-Alexander, COIT. 1 976,
and now 2 1 . ..'it'd8 ! is Geller's sugges­
tion; nevertheless, this may still be a
bit better for White after 22 d6 lLle6 23
exf6 �xf6 24 axb5 axb5 25 �c6.
a2) 13 ....:tb8 (the easiest path) and
now 1 4 lLlc4 �a6 transposes to note
'f' to White's 1 4th move. Instead, 1 4
�b5 lIf8 1 5 �c6 a6 1 6 :l.bl b5 1 7 b4
c4 1 8 �b2 �d7 1 9 �xd7 lLlxd7 = was
0gaard-Tisdall, Norwegian Ch (Kris­
tiansand) 1 987.
b) 13 lIbI ctJh5 ! 14 f4? lLlf6 (Black
should prefer 14 ...i.d4+ ! 1 5 'iii>h l lLlf6
+) 1 5 h3 lIb8 1 6 �d3 a6 1 7 'it'c2 bS 1 8
b3 lLlh5 ! 1 9 axb5 axb5 20 �b2 f5 ! ,
Gulko-Emms, Esbjerg 2000. I f 2 1
exf5 �d4+ 2 2 'iii>h2, then 2 2...'it'h4 +
with the idea 23 fxg6? �xh3 ! '
13 i.a6 (D)
•.•
Black's plan is to eliminate the
powerful c4-knight, and then expand
with moves like ... a6, ... 'it'd7 and ... b5.
202 THE GAMBIT GUIDE TO THE MODERN BENONI
14 i.g5
There are more games with this than
all other moves combined. For exam­
ple, it gets 1 1 pages in Schneider and
everything else gets about a page and a
half. Nevertheless, a couple of White's
alternatives deserve better:
a) Not 1 4 liJa3? due to 1 4. . .liJfxdS !
IS i.xa6 liJxc3 16 'it'd3 liJxa4 - Jano­
sevic.
b) 14 liJe3 i.xe2 I S 'it'xe2 a6 1 6
g4! ? l:tb8 1 7 g S liJhS 1 8 liJg4 bS 1 9
'it'g2 'it'd7 2 0 'lit h 1 ?! bxa4 ! =+= (open
lines) Karasev-Stein, USSR Ch (Le­
ningrad) 1 97 1 .
c) 1 4 i.e3 i.xc4 I S i.xc4 a6 1 6
:el liJd7 1 7 i.fl l:tb8 1 8 'it'c2 bS with
equality, Partos-Matulovic, Bucharest
1 966.
d) 14 i.d2 'it'd7 ! ? I S liJe3 i.xe2 1 6
'it'xe2 a6 1 7 l:tabl b S 1 8 b3 liJhS = 1 9
g 3 fS 20 exfS? 1::tx e3 ! VDanielsen­
Filipowicz, Roskilde 1 978.
e) 14 lIbl gets off the long diago­
nal and prepares b4: 14 ... i.xc4 !? (in­
teresting but unnecessary; 14 . . .'it'd7 is
the standard method, when I S b4 can
be met by I S ...liJhS 1 6 i.d2, and then
1 6 ...fS or 1 6 ...cxb4 1 7 l:txb4 fS) I S
i.xc4 liJd7 ! ? 1 6 liJbS ! liJxbS 1 7 axbS.
Stohl thinks that White has a clear
advantage here, but Black has certain
resources; e.g., 1 7 . . .liJeS ! ? (maybe
1 7 ...i.d4+ 1 8 �h 1 'ifh4 ! ? is a better
choice, since on d4 the bishop exerts a
lot of influence) 1 8 i.e2 gS 1 9 i.e3
( 1 9 'it'd2 h6 20 f4 liJg6) 1 9 ...liJg6 20
'it'd2 h6 2 1 l:tal l:tf8 (considering .. .fS)
22 i.d3 'it'd7 23 l:tfc l liJeS 24 i.fl fS.
White may have something, but his
advantage is not obvious.
f) 14 'lith 1 is particularly notewor­
thy, as it can arise from other move­
orders: 1 4. . .l1b8 I S i.gS !? (again this
move; instead, I S l:.b 1 { to enforce b4 }
IS ...'it'd7 16 b4 is probably best met
by 1 6 . . .liJhS ! =, but 1 6 ... cxb4 17 l:txb4
liJhS 1 8 liJbS i.xbS 1 9 axbS fS ! 20
liJe3 fxe4 2 1 fxe4 liJf6 isn't bad for
Black either because of White's back­
ward e-pawn, Zhidkov-Agzamov, Che­
liabinsk 1 98 1 ) and now:
fl ) IS ...h6 is the conservative line:
16 i.e3 'ile7 (Kapengut; another idea
is 1 6 ...'lith7 1 7 'it'd2 'ild7 !?) 1 7 'it'd2
'lith7?! (better is 1 7 ... i.xc4 1 8 i.xc4
'lith7) 1 8 l:tael (not dangerous; nor is
1 8 b3 liJd7 19 l:tac l fS 20 i.f4 i.xc4 2 1
bxc4 gS, but SeJvanov's 1 8 eS i.xc4
1 9 exd6 'ilxe3 20 'it'xe3 l:txe3 2 1 dxc7
l:tc8 22 i.xc4 l:txc7 23 l:.ael ;!; might
lead one to prefer the 1 6th- or 1 7th­
move options I've indicated; still,
what follows is instructive) 18 ...liJd7
19 i.f2 i.xc4 20 i.xc4 a6 2 1 'ile2 bS !
22 axbS i.xc3 23 bxc3 axbS 24 i.a2 is
"unclear" according to Kapengut. I
think Black is fine here; e.g., 24 ... :a8
2S 'ilc2 :a3 and ...l:tea8.
f2) I S ...'it'd7 and now:
f2 1 ) Interesting is 16 l:.b 1 i.xc4 1 7
i.xc4 a6 1 8 b4 bS 1 9 i.d3 ( 1 9 axbS
liJxbS ! ) 1 9 ... c4 20 i.c2 bxa4 ! 2 1
i.xa4 liJbS with counterplay, Kojder­
Rogulj, Lodz 1 980.
f22) The most critical line is 16 b3,
so that ... i.xc4 can be answered by
bxc4. Then 1 6 . . .liJhS 1 7 l:tc l f6 1 8
i.d2 ! ? ( 1 8 i.e3 f5 1 9 g4 i.xc4 20
bxc4 fxg4 2 1 fxg4 liJf6 = Portisch­
Nunn, London 1 982) 1 8 .. .fS ! 1 9 exf5
gxfS 20 l:.e l occurred in the game
 CLASSICAL MAIN LINE
Ivanchuk-Wahls, Adelaide jr Wch
1 988. Black has excellent activity and
White' s d-pawn is very weak. Instead
of the Wahls's strange 20...f4? ! , I think
that 20 ... �b7 ! is more pointed, espe­
cially with . . .lLlf6 coming.
14 J!Vd7 (D)
••
There is a huge body of theory on
14 . . .h6 1 5 �e3, but I can't find a truly
reliable equalizer for Black, nor against
15 �h4, for that matter. The text­
move unpins the knight, and saves a
tempo in comparison with lines where
the h6-pawn is attacked by a bishop on
e3 and a queen on d2.
15 'iid2
The principal move these days. We
have already seen 1 5 'ii;>h l :ab8 (it
transposes to note 'f2' to White's 1 4th
move). White's other moves resemble
earlier attempts:
a) 15 :el h6 16 �h4 ! ? lLlh5 1 7
�d2 �xc4 1 8 �xc4 a6 1 9 g4 ! ? b5 20
�fl lLlf6 = 2 1 �xf6? �xf6 22 �xh6
b4 23 lLld 1 lLlxd5 24 :c I ? �e7 ! =+
Gyimesi-Shliperman, Guarapuava U-
1 8 Wch 1 995.
203
b) 15 :a3 :ab8 1 6 :b3 :b7 ! ?
(more efficient i s simply 1 6 . . .h 6 1 7
�d2 �h7 1 8 Wh l :e7 = Baragar­
Findlay, Canadian Ch (Ottawa) 1984)
17 lLlb5 �xb5 1 8 axb5 :a8 1 9 �f4
lLlfe8 20 lLle3 :bb8 2 1 'iia l a6 ! 22
bxa6 b5 23 :a3 (23 a7 :b7 24 :a3
lLlf6 25 �a2 lLlce8 26 :al lLlh5 27
�g5 �d4 28 'ii;>h 1 c4 29 lLlc2 �c5 30
b4 �b6 =) 23 . . .:a7 = with the idea
....:.b6 and . . .:bxa6, O.Foisor-Lejeune,
Wattignies U- 1 6 Wch 1 976.
c) 15 ':'bl �xc4 16 �xc4 a6 17 b4
b5 1 8 �d3 c4 ! ? 19 i.c2 bxa4 ! (a
theme to remember: if you don't get
the a-file after axb5 and . . . axb5, then
you can often play . . .bxa4 and win the
b5-square) 20 �xa4 lLlb5 with good
play for Black, Beliavsky-Portisch,
Szirak IZ 1 987.
15...i.xc4 16 �xc4 86 17 'iVd3 (D)
A logical move, stopping ...b5 . The
old main line was 1 7 ':'fel b5 1 8 i.fl
bxa4 ( 1 8 ... h5 ! ? planning ... lLlh7 has
also been played; then Schneider gives
1 9 axb5 axb5 20 'iif4 1L1h7 2 1 �h6 as
t, but 2 l .. .�d4+ 22 �h l b4 23 ':'xa8
Iha8 must be OK for Black) 1 9 ':'xa4
( 1 9 lLlxa4 :eb8 = ) 1 9 . . .1L1fxd5 ! 20
lLlxd5 lLlxd5 2 1 lha6 ':'xa6 22 i.xa6
�d4+ 23 �h 1 lLlc7 24 i.c4 d5 25
exd5 :xel + 26 �xel lL1xd5 =. This is
analysis by Kapengut who, however,
continues 27 �e4(?) lLlb6 =. Schnei­
der points out 27 .. :ii'a4 !, winning at
least a pawn.
After the text-move ( 1 7 "ii'd 3) ,
White has two bishops and a fragile
grip on Black's queenside. His main
problem now is that he's so committed
to stopping . . .b5 that Black has time to
204 THE GAMBIT GUIDE TO THE MODERN BENONI
B (22 Jte2 bxa4 ! ? 23 bxcS �xcS 24
play around on the other side of the
board. Then, too, White's most logical
idea, ':b 1 and b4, often seems to acti­
vate Black's pieces as much as White's.
It's possible that White has a very
small theoretical edge, but even that is
not clear.
17...�h5
Black wants to activate his kingside
pieces, perhaps by . . .JteS and ...�f4,
while watching for an opportunity for
. . .fS.
1 7 . . .'iVc8 ! ? has also been played,
intending ... ':ab8, . . .'iVb7 and . . .bS .
That looks artificial, but isn't so easy
to meet and should definitely be con­
sidered. 1 8 Jtf4 ! .i.f8 1 9 l:.ab1 �d7
and now:
a) Ruban-Filippov, Kemerovo 1995
continued 20 b4 ( ' ? ! ' Stohl) 20 ... cxb4
2 1 .l:.xb4 �eS ! 22 .i.xeS ! ? (not, of
course, 22 .ct2?? as, but Stohl gives
lengthy analysis of 22 'iVe2 bS ! 23
axbS axbS ! with good play for Black)
22 . . .dxeS 23 l:tb2 (23 d6 �e6 ! = in­
tending •cS and . . .�f4), and now
...
easiest was 23 ....i.cS+ 24 'it>h1 'it'd7 ! ?
= with the idea . . .'iVd6, . . ..l:.f8 and . . .fS.
b) Thus Stohl prefers 20 'iVd2 for
White: 20 ...l:.b8 2 1 b4 bS 22 axbS ! ?
lbb8 'iVxb8 2S l:tb1 'iVa7 = ) 22... axbS
(in fact 22 ...�xbS is not bad either;
e.g., 23 .i.xbS axbS 24 bxcS 'iVxcS+ 2S
.i.e3 .i.h6 26 f4 'iVc4 +, due to White's
central weakness) 23 Jte2 c4 24 .i.e3
"with the initiative and the centre", but
this is an instructive misassessment, in
my opinion, because the a-file is a
key factor here and White's e2-bishop
is doing as little as Black's knight on
c7. In the absence of Black's light­
squared bishop, these positions are
generally OK. There might follow
24 . . ..i.g7 2S Jtd4 .i.xd4+ 26 'iVxd4
l:ta8 27 l:ta1 'iib7 (intending ....l:.xa1
and ... l:ta8) 28 f4 ! ? l:txa1 29 .l:.xa1 :ta8
with a good game; e.g., 30 :tel l:.a3 3 1
.i.f3 'iVb6 32 'iVxb6 �xb6 33 ':c 1 f6 !
+. These lines are characteristic of
White's b4 plan in a favourable envi­
ronment; even in that case, Black will
normally achieve good counterplay.
18 g4 (D)
After 1 8 �h1 , instead of 1 8 ... Jtd4?!
19 b4! ;!; Ionov-Scherbakov, Russian
Ch (St Petersburg) 1 998, Scherbakov
recommends 1 8 ... .i.eS as "unclear".
Then 19 f4 .i.g7 ( 1 9 ....i.xc3 20 bxc3
bS is messy) 20 l:tae1 h6 2 1 Jth4
�xf4 ! 22 ':xf4 gS 23 .i.xgS hxgS
looks fine for Black. I should mention
that 1 8 ... h6 also makes more sense in
this case than on move 14.
18 .i.d4+ 19 �hl �g7
••.
We see this manoeuvre in quite a
few Benonis. The knight is itself very
poorly placed, but supports the pawn­
breaks ...hS and ...fS .
 CLASSICAL MAIN LINE
B gives 24 ... g3 2S fS ! , but even this is
20 l:tabl h5! 21 ttJe2
2 1 h3 ! ? is unmentioned; then play
might go 21 ...hxg4 22 hxg4 f6 23 i..h6
i..eS, contemplating moves like ...ttJhS
or ...fS, depending upon the play; for
example, 24 �g2 fS 2S exfS (2S gxfS
gxfS 26 exfS ttJxfS = ) 2S . . . gxfS 26
i..xg7 i.. xg7 27 l:th l ! .:tab8 ! =, with
... bS coming, even after captures on
fS .
21 b5 22 i..a2
.•.
Here Dlugy-de Firmian, USA Ch
(Berkeley) 1988 went 22 . . . i..e S? 23
f4 c4 24 ii'g3 ! ±. Instead, Dlugy offers
22... hxg4 23 ttJxd4 cxd4 24 f4 ! "with
compensation" (24 "ii'xd4 can be met
205
by 24 ...ttJhS ! ? or 24. . . bxa4 = ) . Play
might then continue 24 ...bxa4 (Dlugy
not clear after 2S ... gxh2, intending
...bxa4) 2S 'iVxd4 fS ! ? with the idea 26
eS dxeS 27 fxeS ttJce6. This is cer­
tainly messy enough for both sides,
but I think that Black is fully equal.
The Classical Main Line is not the
threat that it used to be. Reasonably
secure answers have been found, and
White's attempts to restrict his oppo­
nent's position don't seem to have the
staying power that, for example, most
variations of the Modem Main Line
do. In the ...ttJbd7-eS and ...gS lines
(A), Black's piece dance on the dark
squares nullifies White's attempt to
exploit the kingside weaknesses, and a
kind of sustained equilibrium results.
In the ...ttJa6-c7 and ...b6 lines (B),
White has to devote so many forces to
the prevention of ...bS that it's hard for
him to achieve anything positive or
deal with Black on the other side of the
board. If White does manage to keep a
theoretical edge in one or another of
these lines, it tends to be a very small
advantage indeed.
I ndex of Va riations

Chapter Guide 8 'iia4+, 8 i..b5+, 8 i.d3 and 8 h3 -

Chapter 9
1 d4 liJf6 8 liJd2 - Chapter 1 0
2 c4 cS 8 •.• 0-0
3 d5 e6 Now:
4 liJc3 9 i..g 5 - Chapter 3
Other 4th moves - Chapter 1 9 0-0 - Chapter 1 0
4 ... exdS
S cxdS B)
Other 5th moves - Chapter 1 6 e4 g6
S ..• d6 7 f4
Now: 7 liJf3 - see Line A above
A: 6 liJf3 7 g3 and 7 i..b5+ - Chapter 1
B: 6 e4 7 i..f4 - Chapter 2
7 i..g5 - Chapter 3
Or: 7 liJge2 - Chapter 7
6 h3 - Chapter 1 7 f3 - Chapter 8
6 i..f4 - Chapter 2 7 h3 - Chapter 9
6 g3 - Chapter 6 7 i.e2 i..g7 8 liJf3 - Chapter 10 (other
8th moves - Chapter 1 )
A) 7 i.d3 - Chapter 7 (7 . . .i..g7 8 h 3 0-0 9
6 liJf3 g6 liJf3 - Chapter 9)
7 e4 7 ••• i.. g7
7 'iia4+ and 7 i..f4 - Chapter 2 7 ...'iie7 - Chapter 5
7 i..g5 - Chapter 3 8 i.bS+
7 liJd2 - Chapter 4 (with 8 e4 and 9 8 e5 and 8 liJf3 - Chapter 5
i..e2 - Chapter 1 0) 8 .•. liJfd7
7 g3 - Chapter 6 - Chapter 5
7 h3 - Chapter 9
7•.• i.. g7 1 : I ntroduction and Early Moves
7 . . . a6 8 a4 (other moves - Chapter 9) 1 d4 liJf6 2 c4 c5 3 d5 e6 14
8 ... i..g4 - Chapter 9 A: 4 dxe6 15
8 i..e2 B: 4 g3 16
8 i..g5 - Chapter 3 C: 4 liJf3 17
 207

Cl: 4 . . . exd5 1 7 5 cxd5 d6 6 e4 1 7 4: The Knight's Tour Variation

C2: 4 . . . d6 18
D: 4 t2Jc3 1 9 4...exdS 19
D l : 5 t2Jxd5 19
D2: 5 cxd5 20 5 ...d6 6 e4 (6 h3 20)
6...g6 (irregular continuations and trans­
1 d4 t2Jf6 2 c4 c5 3 d5 e6 4 t2Jc3 exd5 5
cxd5 d6 6 t2Jf3 g6 7 t2Jd2 59
A: 7...t2Jbd7 59 8 e4 (8 t2Jc4 60) 8.....tg7
9 t2Jc4 61 (9 i..d3 61 ; 9 ..te2 0-0 10 0-0
l:e8 189) 9 . . t2Jb6 61 (9 . . .'iVe7 61) 1 0
.

positions) 20 t2Je3 62
B : 7 ... ..tg7 64 8 t2Jc4 (8 e4 186) 8 ... 0-0
2: Systems with i.f4 64
1 d4 t2Jf6 2 c4 c5 3 d5 e6 4 t2Jc3 exd5 5 B l : 9 i.f4 65 (9 g3 65) 9 . . .t2Je8 65
cxd5 d6 B2: 9 i.gS 67
A: 6 i.f4 22 B2 1 : 9 ... 'iVd7 ! ? 68 (9 ...t2Ja6 68)
B: 6 t2Jf3 g6 B22: 9 . . :fle7 70
B 1 : 7 'iVa4+!? 24
B2: 7 i.f4 26 5: Pawn-Storm Systems
B2 1 : 7...a6 27 1 d4 t2Jf6 2 c4 c5 3 d5 e6 4 t2Jc3 exd5 5
B21 1 : 8 t2Jd2 27 cxd5 d6 6 e4 g6 7 f4 72
B212: 8 a4 28 A: 7...i.g7 72
B2 1 3 : 8 e4 30 A I : 8 eS 72
B22: 7...i.g7 33 8 'iVa4+ (8 e3 34; 8 h3 A2: 8 t2Jf3 75 8...0-0 9 i.e2 75
34; 8 t2Jd2 34) 8 . . . i.d7 9 'iVb3 34 A2 1 : 9 ... i.g4 76 (9 . . .t2Ja6 75)
C: 6 e4 g6 7 ..tf4 37 7 . . . i.g7 8 'iVa4+ A22: 9 . . .t2Jbd7 79
(8 i.b5+ 37) 8 ... i.d7 9 'iVb3 'ikc7 1 0 A3: 8 i.bS+ 82 8... t2Jfd7 84
t2Jf3 38 A3 1 : 9 t2Jf3 84
A32: 9 i.e2 86
3: Systems with i.gS A33: 9 i.d3 87 9 . . .'iVh4+ ! ? 1 0 g3 88
1 d4 t2Jf6 2 c4 c5 3 d5 e6 4 t2Jc3 exd5 5 1 O. . . 'ilVe7 1 1 t2Jf3 89 1 1 . ..0-0 1 2 0-0
cxd5 d6 t2Jb6 89
A: 6 t2Jf3 g6 7 i.gS 44 7...h6 8 i.h4 44 A34: 9 a4 92 9.. ...lVh4+ (9 . . . a6 93) 10
A I : 8... gS 45 9 i.g3 t2JhS 10 e3 ( 1 0 g3 95 (10 'itfl 95)
'iVa4+ 45; 10 t2Jd2 45) 1 0...t2Jxg3 11 A34 1 : 10 ... 'iVd8 96 11 t2Jf3 0-0 12 0-0
hxg3 i.g7 46 ( 1 2 h3 96) 12...a6 96
A l l : 1 2 t2Jd2 46 A34 l l : 1 3 i.c4 97
A 1 2 : 1 2 i.d3 47 A34 1 2: 13 i.e2 98
A2: 8...i.g7 49 9 e3 (9 t2Jd2 49) 9 . . .0-0 A34 1 3 : 13 i.d3 100
49 A342: 10...'iVe7 101 1 1 t2Jf3 0-0 12 0-0
B : 6 e4 g6 7 t2Jf3 (7 i.g5 50) 7...i.g7 8 a6 102
iLgS 51 A342 1 : 1 3 i.c4 103
C: 6 e4 g6 7 t2Jf3 iLg7 8 i.e2 0-0 9 A3422: 1 3 i.e2 103
iLgS 54 9 . . .h6 54 1 0 iLh4 g5 1 1 i.g3 A3423 : 1 3 i.d3 105
55 1 1 . . .t2Jh5 12 t2Jd2 55 B: 7..:fie7 106
208 THE GAMBIT GUIDE TO THE MODERN BENONI

6: Fianchetto Systems with g3 B: S lDge2 150

1 d4 lDf6 2 c4 c5 3 d5 e6 4 lDc3 exd5 5 C: S i..e3 152 S...O-O 152
cxd5 d6 6 lDf3 (6 g3 110) 6 g6 7 g3
.•. C l : 9 lDge2 153
110 7 ... i..g7 S i..g2 0-0 9 0-0 110 (9 C2: 9 'iWd2 154
lDd2 1 1 1 ) D: S i..g5 155
A : 9...a6 1 0 a4 lDbd7 1 12
A I : 1 1 e4 113 9: Modern Main Line
A2: 11 i..f4 114 1 d4 lDf6 2 c4 c5 3 d5 e6 4 lDc3 exd5 5
A3: I1 lDd2 115 cxd5 d6 6 lDf3 g6
A3 1 : I1 ..JIeS 115 A: 7 h3 157 7 ... a6 8 a4 'fie7 157
A3 1 1 : 1 2 lDc4 1 1 7 B : 7 e4 161
A3 1 2: 1 2 h 3 118 B l : 7...a6 161 (7 . . .i.. g4? 161)
A32: 1 1 ...lDh5 123 B l l : S a4 1 62 S...i..g4 162
B: Lines with ...lIeS 124 B I l l : 9 h3 162
B l : 9 . . .:e8 124 B 1 1 2 : 9 i..e2 164
B2: 9 . . .a6 10 a4 :e8 127 B I 1 3 : 9 'ilVb3 1 68
B 1 2: S h3 1 71 8 . . . b5 9 i..d3 i..g7 1 0
7: Systems with i..d 3 and lDge2 0-0 1 71
1 d4 lDf6 2 c4 c5 3 d5 e6 4 lDc3 exd5 5 B2: 7 ... i..g7 1 76 S h3 0-0 9 i..d3 1 77
cxd5 d6 6 e4 g6 7 i..d3 130 (7 lDge2 9... lDh5!? 10 0-0 lDd7 1 79
i.. g7 8 lDg3 130) 7...i..g7 (7 . . . a6 132) B2 1 : 1 1 i..g5 1 79
S lDge2 132 (8 h3 132) S...O-O 132 9 B22: 1 1 i..e3 182
0-0 132 B23: I I l:!e1 183
A: 9...a6 134 (9 ...:e8 134; 9 . . . lDbd7
135) 10 a4 lDbd7 11 h3 'fic7 12 lDg3 10: Classical Main Line
:eS 13 f4!? c4 14 i.. c2 :bS 15 i..e3 1 d4 lDf6 2 c4 c5 3 d5 e6 4 lDc3 exd5 5
lDc5 137 cxd5 d6 6 lDf3 g6 7 e4 186 (7 lDd2
A I : 16 'fif3 138 i..g7 8 e4 186) 7 ... i..g7 (7 . . . a6 186) S
A2: 16 i..d4 140 i..e2 186 (8 lDd2 186) S...O-O 9 0-0
B : 9...lDg4 142 186 9 ...:eS!? (9 . . . a6! 10 a4 i..g 4!
187) 10 lDd2 188 ( 1 0 'fic2 188)
8: Kapengut's 7 f3 System A: 10 ... lDbd7 189
1 d4 lDf6 2 c4 c5 3 d5 e6 4 lDc3 exd5 5 A I : 1 1 h3 190
cxd5 d6 6 e4 g6 7 f3 1477...i..g7 147 A2: 1 1 a4 194
A: S i..d3 148 B: 10 ...lDa6 199 1 1 f3 199