Learn why your pawn structure decides whether your bishop is a long-range attacker or a tall pawn.
✓ Interactive boards ✓ Step-by-step ✓ Free forever
A good bishop is one whose own pawns sit on the opposite color, leaving its diagonals open and its targets reachable. A bad bishop is the reverse: its own pawns clog the squares it travels on, turning a long-range piece into a tall pawn. The classic example is the French Defense, where Black's light-squared bishop is buried behind pawns on d5, e6 and f7 — it sees nothing and goes nowhere. Wilhelm Steinitz wrote about this 140 years ago and it has been the engine of strategic chess ever since: a good bishop and a bad bishop is one of the largest static imbalances in the game.
The good bishop / bad bishop distinction was formalised by Wilhelm Steinitz in the 1880s and became one of the cornerstones of Siegbert Tarrasch's positional rules. The idea is simple: bishops travel on a single color, so the value of a bishop depends entirely on which color squares your pawns occupy. José Raúl Capablanca built half his world-championship style around exchanging into endgames where his bishop was on the opposite color from his pawn structure. Decades later, Bobby Fischer used the same principle to crush Tigran Petrosian in the 1971 Candidates Final and Boris Spassky in their 1972 World Championship match. The lesson has not changed in 140 years: place your pawns on the color opposite your remaining bishop, and you will win endgames you should have drawn.
Each bishop only ever moves on one color. The light-squared bishop starts on f1 (White) or c8 (Black) and stays on light squares forever. The dark-squared bishop is its mirror. Because the color is permanent, the only variable is your pawn structure — and that is what makes a bishop good or bad.
If many of your own pawns sit on light squares and you still have your light-squared bishop, those pawns are blocking the bishop's diagonals from the inside. The bishop has nowhere to go. Worse, the squares the pawns DON'T cover are also unreachable for the bishop — meaning your weak squares stay weak and your active squares are unused.
If your pawns sit on dark squares, your light-squared bishop is unblocked and free to roam. Even better, the pawns and bishop now cover both colors as a team: pawns on dark squares + bishop on light squares = full board coverage. This is the structural setup every endgame technician aims for.
Black's light-squared bishop on c8 is buried by pawns on d5, e6 and f7. The French Defense is built around how to fix or trade this piece.
White's bishop on d3 is the textbook good bishop: pawns on dark squares (e4, d4) leave the light diagonals fully open.
Black plays ...c4 (or ...cxd4) to change the structure. The c-pawn moves off light squares and the diagonal a4-e8 opens for the bad bishop.
Black's bishop on a4 is structurally still 'bad', but it pins the c3-knight and pressures the queenside. Bad on paper, useful in practice.
A bishop is bad because it is on a passive square
A bishop is bad because of YOUR pawn structure, not because of the square it sits on. The light-squared bishop in the French is bad even on c8, even on b7, even on h3 — its problem is that Black's d5, e6 and f7 pawns block every useful diagonal. Move the bishop and the problem moves with it.
A bad bishop is always worse than a knight
A bad bishop OUTSIDE the pawn chain is often very useful — Mihail Suba called this the 'active bad bishop'. The classic French move ...Bd7-a4 puts the 'bad' bishop on a square where it cuts the queenside in half. The bishop is still 'bad' by Steinitz's definition (its pawns block its diagonals) but it is doing real work. 'Bad' is structural; 'passive' is functional. They are not the same thing.
If I have a bad bishop, I should always trade it for a knight
Not always. If your opponent's pieces and pawns are all on the bishop's color, your bad bishop suddenly becomes the only piece that can attack their weaknesses. Mark Dvoretsky called this the 'bad bishop that defends good pawns' — it is bad on offense, but invaluable on defense. Read the position before you trade.
Test yourself with these positions
It is Black to move in this French Defense Advance Variation. Both sides still have both bishops. Which of Black's bishops is bad, and why?
It is Black to move. Black's light-squared bishop is locked in by the e6 pawn. Find the pawn break that frees the bishop.
Black's light-squared bishop is on a4 — outside the pawn chain. Is this bishop good, bad, or 'active bad'?
Find the move that exploits — or fixes — the bishop imbalance
Black to move in the French Advance. The light-squared bishop is bad. Find the move that opens the c-file and gives the bad bishop a route off the back rank.
White to move. Find the move that wins material AND simplifies into a structurally favorable endgame for the good-bishop side.
White to move. Black's dark-squared bishop on g7 is restricted by pawns on d6 and e7-f7-g6. Find the knight maneuver that targets the cornerstone of that bad-bishop structure.
These openings are decided by which side gets the better bishop
The French is the textbook good-bishop / bad-bishop opening. Black's c8-bishop is the most famous bad bishop in chess — buried behind pawns on d5, e6 and f7. Every French line revolves around how Black can either trade it (the ...Bd7-e8-h5 maneuver) or activate it outside the chain (...Bd7-a4). Knowing this concept is the difference between playing the French well and being slowly squeezed.
View opening pageThe Caro-Kann is the French's solution to the bad-bishop problem. By playing 1...c6 first, Black prepares ...d5 AND keeps the c8-bishop's diagonal open. The price is a slower setup, but the bishop emerges to f5 or g4 — usually an active and useful piece, not a problem to solve.
View opening pageIn the Queen's Gambit Declined, Black accepts a slightly restricted light-squared bishop in exchange for a solid central pawn structure. The maneuver ...b6 and ...Bb7 (the Tartakower variation) is a direct attempt to give the bad bishop a useful diagonal — at the cost of a tempo. Fischer-Spassky 1972 Game 6 is the classic illustration of how this structure decides middlegames.
View opening pageFischer's seventh-game win against Petrosian is one of the cleanest demonstrations of good-bishop / bad-bishop technique ever played at the world-championship level. Fischer fixed Petrosian's pawns on dark squares, traded into an endgame with his good light-squared bishop versus Petrosian's restricted dark-squared bishop, and converted the structural advantage with mechanical accuracy. The game is annotated in dozens of strategy books.
Fischer's QGD masterpiece. He developed his dark-squared bishop to a free diagonal while Spassky's was blocked by his own pawns. The middlegame became a model good-vs-bad-bishop position, and Fischer converted with the precise minority attack he had studied for years. Spassky reportedly applauded at the end — one of the few moments of public sportsmanship from the most famous chess match in history.
The textbook good-bishop endgame. Capablanca traded into a rook-and-bishop ending where his bishop was on light squares and Tartakower's pawns were also fixed on light squares — meaning Tartakower's bishop had nothing to do, and his pawns were targets for Capablanca's bishop. Capablanca then walked his king up the board, winning a position that looked drawish to most strong players. The game taught a generation what 'technique' meant.
Pitfalls in handling the bishop imbalance
1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 Be7 5.e5 Nfd7 6.Bxe7 Qxe7
After 6.Bxe7 Qxe7, White has eliminated his GOOD dark-squared bishop, while Black has kept his BAD light-squared bishop on c8 — but more importantly, Black's bishop is now the only minor piece that can defend the d5 pawn structure. The exchange is exactly backwards from the structural point of view. White should keep his good bishop and force Black to deal with the bad one.
1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 Be7 5.Bg5 h6 6.Bh4 O-O 7.e3 b6 8.Bd3 Bb7 9.O-O Nbd7 10.Qc2 c5 11.cxd5 exd5
After 11...exd5, Black has 'recaptured the right way' (keeping the e-file half-open) but now the d5 pawn locks in the c8-bishop's natural diagonal. Black's b7-bishop, on the long light diagonal, must do double duty — defend c6 AND attack along a8-h1. Many club players reach this structure without realising they have committed to a permanently restricted light-squared bishop.
1.e4 e6 2.d4 d5 3.Nd2 c5 4.exd5 exd5 5.Ngf3 Nc6 6.Bb5 Bd6 7.O-O Nge7 8.dxc5 Bxc5 9.Nb3 Bd6 10.Bg5 O-O 11.Bh4 Bg4
Black's ...Bg4 places the 'bad' light-squared bishop OUTSIDE the chain, where it pins the f3-knight. White suddenly faces a fully active 'bad' bishop. The trap for White: thinking the bishop is bad meant 'no longer dangerous'. Suba's rule applies — apply the structural definition, but do not stop there.
After the opening, count your bishops and check what color squares your pawns sit on. Pawns on the SAME color as your bishop = bad bishop. Pawns on the OPPOSITE color = good bishop.
If your bishop is bad and you cannot fix the pawn structure, plan a trade. ...Bd7-e8-h5 in the French is the textbook maneuver.
If your bishop is good, refuse trades. A good bishop in an endgame beats a bad bishop or a knight almost automatically.
Pawn breaks are bishop fixes. The French ...c5 break and the QGD ...c5 break exist mainly to free a bad bishop.
An 'active bad' bishop OUTSIDE the pawn chain (like ...Ba4 in the French) can do real work. Don't auto-trade it just because it's structurally bad.
When trading, look at YOUR opponent's pawn structure too. Trading their good bishop for your bad bishop wins the structural battle every time.
Everything you need to know about good and bad bishops
A good bishop is one whose own pawns sit on the opposite color, leaving its diagonals open. A bad bishop is one whose own pawns sit on the same color, blocking its movement. The bishop's color is fixed by its starting square, so 'good' and 'bad' depend entirely on your pawn structure. The classic example is the French Defense, where Black's pawns on d5, e6 and f7 trap the light-squared bishop — making it the textbook bad bishop.
There are three standard methods: (1) play a pawn break (like ...c5 or ...f6) that moves your worst pawn off the bishop's color; (2) trade the bad bishop using a maneuver such as the famous French ...Bd7-e8-h5 route to swap it for the opponent's good bishop; or (3) reroute the bishop OUTSIDE the pawn chain — the so-called 'active bad bishop' — where it can do real work even though the structure is unchanged.
No. A bad bishop INSIDE the pawn chain is usually worse than a knight, but a bad bishop OUTSIDE the pawn chain is often equal or better. Mihail Suba called this the 'active bad bishop'. The structural definition tells you about the pawn chain; it does not tell you whether the piece is doing useful work. Always check both.
Because a structural advantage in an endgame is one of the most reliable wins in chess. If you trade your bad bishop for the opponent's good bishop, you walk into the endgame with the better piece on the board. Capablanca, Petrosian and Karpov built their careers on this single trick: simplify into endgames where your bishop is good and the opponent's is bad, then convert with technique.
Yes. Kingsights analyzes your games and identifies positions where you played with a structurally bad bishop (or your opponent did), and whether you fixed it, traded it, or let it suffer. Enter your Chess.com username above to see how often the good-bishop / bad-bishop battle decided your games.
Kingsights scans your real games to find positions where the bishop battle decided the result.
✓ Interactive boards ✓ Step-by-step ✓ Free forever