Trap · played as White

Caro-Kann: Réti–Tartakower Trap

After 4...Nf6 5.Qd3, Black's premature 5...e5?? runs into 8.O-O-O and the crushing 9.Qd8+!! queen sacrifice mating by force, and the more common 5...Bf5? drops a piece to 6.Nxf6+ and 7.Qxf5; declining with 5...Nxe4 keeps Black safe but hands White an easy game.

A position this opening is aiming for.

You play
White
Variations covered
15
Branch points analysed
3
Club games studied
62k
After 6 moves
close to winning for you
What it wins
Checkmate

The main line

This is the line you will reach most often, and the one everything else branches away from. You are White; every move here has been checked with an engine so that nothing in it drops material.

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Qd3 e5 6.dxe5 Qa5+ 7.Bd2 Qxe5 8.O-O-O Nxe4

The decisions you actually face

These are some of the positions where Black has a genuine choice, and this is what Black really chooses at club level. Every percentage below is measured from Lichess games rated 1600 to 2499, not from what a book says people ought to play. Of the 11 replies you will meet most often here, this repertoire answers 8.

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Qd3

62k games · Black to move

They playHow oftenThey scoreYou answer
Nxe451%50%Qxe4
Nbd723%52%Nf3
Bf56%20%Nxf6+
g65%56%Nf3

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Qd3 e5 6.dxe5

450 games · Black to move

They playHow oftenThey scoreYou answer
Qxd355%42%Bxd3
Qa5+32%25%Bd2
Nxe411%29%Qxe4

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Qd3 Nxe4 6.Qxe4

76k games · Black to move

They playHow oftenThey scoreYou answer
Nd760%52%Bc4
Qd514%53%
g613%49%
e69%39%

Frequencies and scores from the Lichess opening explorer, 2026-09-14. An empty cell means this repertoire does not prepare that move, usually because it is rare enough to work out over the board.

The mistake to watch for

One move in this opening is both popular and bad, which is the combination worth knowing. After 1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Qd3, Black plays Bf5 in 6% of club games, and it costs 4.3 pawns against best play. The best reply is Nxe4.

The position before the mistake. Black to move.

How you punish it

Bf5 Nxf6+

Verified with Stockfish at depth 26. That matters more than it sounds: a shallow search will call all sorts of normal moves refutations, and this one was re-checked deeper before it was written down.

Can you actually reach it?

Not from any repertoire in this library. That is worth saying plainly, because most trap collections never check. Springing it needs every White move in 1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 to be a move your book already prepares, and none of the White repertoires here carry prepared moves that far down this particular line. Often the opening is the same and the split comes later, when your opponent plays something your repertoire answers a different way. It is still worth knowing from the other side of the board: this is a position where the natural move loses.

How the trap springs

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Qd3 e5 6.dxe5 Qa5+ 7.Bd2 Qxe5 8.O-O-O Nxe4 9.Qd8+ Kxd8 10.Bg5+ Kc7

Traps are only worth knowing if somebody actually falls into them, so this one was found by cross-referencing what club players really play against what an engine says it costs, rather than by picking a pretty line out of a book. Black has to go wrong in a way Black genuinely does go wrong.

Some of the variations

…and 9 more in the trainer.

Train it

ChessReflex drills the branch points where a line splits rather than making you replay it move by move. Free, and no account needed to start.

Open Caro-Kann: Réti–Tartakower Trap