Nimzo-Indian Defence: the Rubinstein Variation and the ...Ba5 trap
4.e3 is White's most played plan against the Nimzo-Indian, and the main tabiya measures close to equal. The line where Black keeps the bishop with ...Ba5 hides a concrete trap we measured with the engine: only one move recaptures correctly on c3, and the rest loses material.
5 min read · Updated on 2026-09-12
Short answer
The Rubinstein Variation arises from 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3, White's soundest and most played plan against the Nimzo-Indian. The main tabiya, after the trade on c3, measures close to equal — the Nimzo-Indian is rightly respected as Black's most reliable defence against 1.d4. But the line where Black avoids the trade with ...Ba5 hides a tactical detail this guide measures precisely: a gap of over two pawns between the right move and the wrong one in the same position.
The essentials in a few lines
- After 4.e3 O-O 5.Bd3 d5 6.Nf3 c5 7.O-O Nc6 8.a3 Bxc3 9.bxc3 dxc4 10.Bxc4, the position measures between +0.16 and +0.48 for White depending on Black's reply — the Nimzo-Indian gives Black one of the most balanced positions against 1.d4.
- In the line where Black keeps the bishop with 8...Ba5, only ...Bxc3 recaptures correctly after 12.dxc5: it measures +0.19, against +2.70 (...Bc7) and +2.95 (...Bg4) for the engine's other options — a gap of over two pawns for the same general idea.
- The concrete reason behind ...Bg4's failure is a pawn move: 11.b4! attacks the bishop on a5 and the engine measures the resulting position at +2.83, because the bishop has no good retreat after the advance.
- After ...Bxc3, White must recapture immediately with 12.bxc3 (+0.26); delaying the recapture with 12.Rb1 or 12.Bc2 costs between 5.08 and 5.50 pawns of evaluation — a mistake that changes the game's outcome, not just its evaluation.
- The Rubinstein Variation rewards specific tactical precision as much as positional understanding: the ...Ba5 line looks more passive than the direct trade on c3, but hides the largest margin for error measured in this guide.
The main tabiya after the trade on c3
Who stands better after Black trades the bishop on c3?
White, with a small and stable edge between +0.16 and +0.48 depending on Black's reply — the Nimzo-Indian is respected precisely because, even giving up the bishop pair, Black reaches sound positions without a clear disadvantage.
After 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 O-O 5.Bd3 d5 6.Nf3 c5 7.O-O Nc6 8.a3 Bxc3 9.bxc3 dxc4 10.Bxc4, White has recovered the c4-pawn and holds the bishop pair against doubled c3-pawns. This is the classical Rubinstein structure: White plays for the bishop pair and the central e4 advance, Black plays against the doubled pawns with ...b6 and ...Ba6 or ...Bd7-a4.
The three black replies closest in evaluation — 10...Qc7, 10...h6 and 10...b6 — measure +0.28, +0.45 and +0.48. The difference between them is about which piece Black prioritises developing first, not about the position's safety.
Every evaluation in this and the following section comes from Stockfish 19 running locally, depth 22, measured by us from the stated lines.
The ...Ba5 line and the 11.b4 trap
Why is 8...Ba5 more dangerous for Black than it looks?
Because it delays the trade on c3 without removing the threat: after 9.cxd5 exd5 10.dxc5, if Black retreats the bishop with ...Bc7 or ...Bg4 instead of capturing with ...Bxc3, White plays 11.b4! and the bishop has no good retreat — the engine measures that position at +2.83.
Keeping the bishop with 8...Ba5 instead of trading immediately looks more ambitious: Black preserves the bishop and keeps pressure on the c3-knight. The problem appears three moves later, when White captures the c5-pawn with 10.dxc5.
In that position, Black's only correct recapture is 10...Bxc3, which the engine measures at +0.20 — practically equal. Any attempt to keep the bishop active with 10...Bc7 or 10...Bg4 costs far more: both measure above +2.4, because neither resolves where the a5-bishop ends up.
The concrete reason is a simple pawn move: after 10...Bg4, 11.b4! attacks the bishop directly, and it has no good retreat square — a5 and b6 are covered, and any retreat loses time and leaves White up a pawn with the initiative. The resulting position measures +2.83, nearly three pawns of White advantage for the same general idea that, one move earlier, was worth +0.19.
| ♜ | ♝ | ♛ | ♜ | ♚ | |||
| ♟ | ♟ | ♟ | ♟ | ♟ | |||
| ♞ | ♞ | ||||||
| ♝ | ♙ | ♟ | |||||
| ♙ | ♘ | ♗ | ♙ | ♘ | |||
| ♙ | ♙ | ♙ | ♙ | ||||
| ♖ | ♗ | ♕ | ♖ | ♔ |
Full line: 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 O-O 5.Bd3 d5 6.Nf3 c5 7.O-O Nc6 8.a3 Ba5 9.cxd5 exd5 10.dxc5
After ...Bxc3, the recapture cannot wait
What happens if White does not recapture on c3 immediately?
White loses the game in practice, not just in evaluation: after 10...Bxc3, 11.bxc3 measures +0.26 and keeps the position balanced, but delaying the recapture with 11.Rb1 or 11.Bc2 costs between 5.08 and 5.50 pawns — Black ends up with a decisive advantage.
Once Black correctly captures with ...Bxc3, the knight (or what is left of it) sits on c3 as a captured piece — recapturing is the absolute priority, because that piece is what keeps White's pawn structure coherent and what opens the possibility of b4 later.
11.bxc3 resolves this immediately and measures +0.26 — the position returns to the same kind of slightly White-favourable balance seen in the Rubinstein main line. Any attempt to play something else first, such as 11.Rb1 (−5.08) or 11.Bc2 (−5.50), leaves Black with an advantage no future positional compensation reverses.
For anyone studying this repertoire: the lesson is not "memorise the ...Ba5 line", it is understanding that it has a single correct answer at two concrete points — Black's recapture on c3 after dxc5, and White's recapture after ...Bxc3. Outside those two points, the game follows the general Rubinstein structure plans.
Frequently asked questions
Does the Rubinstein Variation give White a large advantage?
No. The main tabiya measures between +0.16 and +0.48, a normal opening edge — the Nimzo-Indian remains one of the most reliable defences against 1.d4.
Why is 8...Ba5 riskier than it looks?
Because after 10.dxc5, only ...Bxc3 recaptures correctly (+0.20); the alternatives ...Bc7 and ...Bg4 measure above +2.4, because the a5-bishop has no good retreat after 11.b4.
What does White need to remember after ...Bxc3?
Recapture immediately with bxc3. Delaying the recapture costs between 5.08 and 5.50 pawns of evaluation — the piece on c3 cannot be left there.
Where do this guide's evaluations come from?
From Stockfish 19 running locally, depth 22, measured by us from the stated lines — not from a third-party opening database.
Openings mentioned in this guide
- Nimzo-Indian Defense (E20)
- Queen's Indian Defense (E12)
- Bogo-Indian Defense (E11)
See if you recapture on the right move
Import your Nimzo-Indian games and see whether recapture mistakes in simple tactical positions are costing you evaluation that theory has already resolved.
Create a free account