How to Play Against the Nimzo-Indian Defense
3...Bb4 is a direct structural threat, not a developing move played for its own sake. Black is offering to trade the bishop for your c3-knight specifically to damage your queenside pawns, betting that the resulting weaknesses will matter more over the course of a game than the bishop pair and extra central space you get in return. How you answer that threat — accept the trade, avoid it, or force it on your own terms — shapes the entire rest of the opening.
The solid main line: the Rubinstein System
Playing 4.e3 develops naturally and doesn't try to prevent the trade at all — if Black plays ...Bxc3+, you recapture with bxc3, accept the doubled pawns, and rely on the bishop pair and extra central influence to compensate over time. This is the most common, most reliable choice, and it's sound at every level. The one thing it requires that many club players skip: an actual plan for using the resulting structure, typically central expansion with e4 or targeted pressure down the newly opened b-file, rather than just accepting the trade and playing generic developing moves afterward.
Avoiding the trade entirely: 4.Qc2
The Classical Variation sidesteps the whole structural question. If Black takes on c3, you recapture with the queen instead of a pawn, keeping your structure completely intact at the cost of a slightly less developed kingside for a few moves. This has been a serious main line at the highest level for exactly that reason — it removes Black's main structural weapon from the position entirely, forcing the game to be decided on piece play and central control rather than a long-term pawn-structure argument.
The aggressive option: forcing the issue with the Sämisch
Rather than waiting for Black to decide when to trade, 4.a3 forces the question immediately — Black almost always plays 4...Bxc3+ 5.bxc3, and White gets the bishop pair and a strong pawn center on move five instead of move eight or nine. The Sämisch Variation commits to using that structure actively: quick development, f3 and e4, and a central pawn storm rather than a slow buildup. It's a genuinely sharper try than the Rubinstein, with real winning chances if the center opens in your favor, but it also commits your structure earlier, giving Black more certainty about what to prepare against.
The mistake that wastes the whole trade
Whichever structure you accept, playing passively afterward squanders the compensation for doubled pawns — developing pieces to reasonable squares without ever actually using the bishop pair, the extra central space, or the open b-file to create real threats. The doubled c-pawns are a genuine long-term weakness that Black will happily target for 40 moves if you let the game drift into a quiet, symmetrical-feeling middlegame. Your entire justification for accepting the structural damage in the first place is dynamic, genuinely active play; without it, you've simply accepted a weaker pawn structure for nothing in return.
Watch for Black avoiding the pin altogether
Not every 1...Nf6 2.c4 e6 player commits to 3...Bb4. Some play 3...d5 instead, transposing toward Queen's Gambit Declined structures, or 3...b6, heading for a Queen's Indian setup with a fianchetto instead of a pin. Neither is a mistake on Black's part — they're simply different openings that happen to share a move order for two moves. Don't assume Nimzo-Indian-specific preparation still applies once Black's third move signals a different structural plan.
The Leningrad idea: pinning back before Black pins you
A fourth, less common try, 4.Bg5, pins Black's own knight on f6 rather than reacting to the threat on c3 at all. It's a sharper, more tactical alternative that keeps the tension over the c3 knight unresolved while creating an immediate problem of Black's own to deal with. It sees less club-level play than the three main systems above mostly because it demands more precise follow-up knowledge, but it's a legitimate way to unbalance the position early if the Rubinstein's slower structural battle doesn't appeal to you.
Have a plan for the structure before you accept it
Doubled pawns aren't automatically compensated by the bishop pair — they're compensated by what you do with the bishop pair. Decide your active follow-up plan before you play 4.e3 or 4.a3, not after Black has already traded on c3.
For the closely related setup one move earlier, see how to play against the Bogo-Indian Defense, which appears when White avoids 3.Nc3 entirely.
Frequently Asked Questions
Is the Rubinstein System's doubled c-pawns really worth the bishop pair?
At club level, yes, in most practical games — the bishop pair and central space tend to matter more than the structural weakness before an endgame is reached, and most club games are decided well before a pure pawn-structure endgame becomes relevant. At higher levels the balance is closer and depends heavily on precise play from both sides.
Which system should a player with limited study time actually pick?
4.e3 (Rubinstein) is the most forgiving — it develops naturally, doesn't require memorizing forcing lines, and the underlying plan (central expansion, use the bishop pair) is easy to reconstruct from general principles if you forget the exact move order.
Does 4.Nf3 avoid all of this by developing before deciding?
It delays the decision rather than avoiding it — Black can still play ...Bxc3+ later, or the game can transpose into one of the systems above depending on subsequent moves. Treat 4.Nf3 as a flexible move order, not a fourth independent plan.