You reach an endgame a clean pawn ahead. Maybe two. Your opponent has nothing — just a lone bishop shuffling back and forth. You start pushing your passers, certain the point is in the bag… and forty moves later you are shaking hands on a draw, wondering what on earth just happened.
If that scenario makes you wince, you have met the most misunderstood ending in club chess: the opposite-colored bishop (OCB) endgame. It breaks the normal arithmetic of chess. “I’m up material, therefore I’m winning” is a rule that simply stops applying here, and adult improvers lose half-points to it every single week. This guide explains why the extra pawn evaporates, exactly how much material you actually need to win, and — the part almost nobody teaches — the moment several moves earlier where the game was really decided.
Why your extra pawn just disappears
In a normal bishop endgame, the side that is up a pawn wins because the bishops can challenge each other, trades happen, and material talks. Opposite-colored bishops can never do that. Your bishop travels on light squares; theirs lives on dark squares. They will never meet, never trade, never contest the same square. The result is that each bishop is effectively a permanent, untouchable guardian of one color complex.
That single fact rewrites the endgame. The defender does not need to match your material — they only need to control the squares your pawns must cross. Because your bishop can never evict their bishop from those squares, a single well-placed enemy bishop can hold back two or even three pawns by itself. The extra pawn is still on the board; it just has nothing to do.
The drawing margin: how much do you actually need?
Think of every OCB endgame as having a “drawing margin” — the amount of extra material the defender can absorb before the position genuinely tips. For most other endings that margin is razor-thin. In OCB endings it is enormous. Here is the practical hierarchy I teach, from “almost always a draw” to “real winning chances.”
One pawn up — assume it’s a draw
A single extra pawn is, with very few exceptions, not enough. The defender parks their bishop on the diagonal in front of the pawn and brings the king alongside. There is no second front to stretch them, so they never run out of moves. Don’t grind for twenty moves hoping it cracks; it usually won’t.
Two connected passed pawns — still probably a draw
This is the trap that fools strong club players. Two connected passers look overwhelming, but if they can be blockaded before they reach the sixth rank, the defending bishop sits in front of them on its own color and the king covers the other square. Nothing gets through. Connected pawns are only decisive when they are already far advanced and supported by your king at the moment of contact.
Separated passed pawns — now we’re talking
The genuine winning structure is passed pawns on opposite wings, far enough apart that one bishop cannot watch both. The wider the gap — ideally two, three, or more files — the better. The idea is brutal and simple: the bishop blockades one pawn, your king escorts the other, and the defender is one piece short of covering both fronts. This is the “two-weaknesses” principle in its purest form.
The defender’s job: build the blockade
If you are the one down material in an OCB ending, your survival plan is almost mechanical, and knowing it is worth a lot of saved half-points:
1. Put your bishop on the right color. Your bishop should blockade the most dangerous passed pawn from in front, on the diagonal the pawn cannot leave. 2. Use your king for the other color. The squares your bishop can’t touch are exactly the ones your king must guard. 3. Refuse to be stretched. The only way you lose is a successful second front, so keep both your king and bishop within reach of both wings. Build the fortress, then make waiting moves forever.
Defenders draw these endings not by calculating brilliantly but by recognizing early that the fortress exists and steering straight for it. That recognition is the same skill you build when you study key squares and the opposition in king-and-pawn endgames — it’s about controlling squares, not counting wood.
The attacker’s job: two fronts, then the breakthrough
If you are trying to win, hope is not a plan. You need a method:
First, create a second passed pawn as far from the first as possible. A pawn majority on the opposite wing is gold here even if it means conceding material elsewhere — separation matters more than the pawn count. Second, advance your pawns onto squares your opponent’s bishop does not control. Pushing a pawn onto the enemy bishop’s color just hands them a free blockade. Third, bring your king to do the heavy lifting on the color your own bishop can’t cover. And finally, be ready to sacrifice the bishop at the moment one pawn promotes — giving up your last piece to make a queen is the standard winning trick, not a desperation move.
The mistake that happens five moves before the endgame
Here is the counterintuitive rule that quietly costs adult improvers the most points, and almost no beginner guide mentions it:
Opposite-colored bishops are drawish in pure endgames but favor the attacker when other pieces are still on the board.
With rooks or queens still on, your extra bishop becomes an attacker with no defensive counterpart — it’s like playing with an extra piece on its color complex around the enemy king. So the worst thing you can do when you’re up material and pressing an attack is to “simplify into a winning endgame.” Trade everything off and you walk straight from a near-winning middlegame into a dead-drawn OCB ending. The position didn’t betray you on move 50; you traded the win away on move 35.
This is exactly why a reliable trading framework matters. Before you swap that last rook, ask what structure you are heading into — the habit is the same one covered in our decision framework for when to trade pieces. With opposite-colored bishops the answer flips the usual logic: when you’re ahead, keep the major pieces on.
A four-question test before you simplify
Run this quick check any time an OCB structure is about to appear on the board:
1. Will any other pieces remain? If yes and you’re attacking, keep them on — OCB favors you here. 2. Are my passed pawns connected or separated? Separated and wide is winning; connected and blockable is usually a draw. 3. Can the enemy bishop reach a square in front of my pawns? If one bishop can blockade everything, the draw is in. 4. Can my king cover the squares my bishop can’t? If not, you may be the one who has to hold. Four questions, ten seconds, and you’ll stop converting won games into draws.
How to actually train this
Reading about OCB endings is not the same as feeling them over the board. The fastest way to internalize the drawing margin is to review your own games where you let a won endgame slip and tag every position where opposite-colored bishops appeared. You will almost certainly find a pattern: a premature trade, a pawn pushed onto the wrong color, or a missed chance to keep the rooks on. Patterns you find in your own losses stick far better than any memorized rule.
From there, set up the three model positions — one pawn up, two connected pawns, separated passers — against an engine on the defending side and try to hold the draws. Once you can defend them, you’ll instantly recognize when you’re attacking one, and that recognition is worth more than any single line of theory.
Find out which endgames your style leaks points in
Opposite-colored bishop endings are just one place where adult improvers bleed half-points — and where you leak the most depends on the kind of player you are. Take the free assessment and get your personalized chess archetype report, which pinpoints the exact phases costing you rating.
If you’d rather skip the guesswork entirely, the MyChessPlan personalized study plan ($14.99) turns that assessment into a week-by-week training roadmap — including the endgame patterns most relevant to your games — so you spend your study time where it actually moves your rating.
Frequently asked questions
Are opposite-colored bishop endgames always a draw?
No — but the drawing tendency is very strong. A single extra pawn is almost always a draw, and even two connected passed pawns frequently draw if they can be blockaded. The reliable winning structure is passed pawns on opposite wings, far enough apart that one bishop cannot stop both while your king escorts the other.
Why are opposite-colored bishops good for the attacker in the middlegame but drawish in the endgame?
With rooks or queens still on the board, your bishop acts as an extra attacker on its color complex with no enemy counterpart, which intensifies an attack. Once everything else is traded off, that same bishop can only ever guard one color, so a lone defending bishop on the other color builds an unbreakable blockade. The pieces flip the verdict.
How many extra pawns do I need to win an opposite-colored bishop endgame?
There’s no fixed number — separation matters more than count. Two connected pawns can draw, while two widely separated passers usually win. Aim to create passed pawns on opposite wings with at least two or three files between them, then use your king on the squares your bishop can’t cover.
I’m defending and down a pawn — what’s my plan?
Build a fortress. Blockade the most dangerous pawn with your bishop from in front, guard the other color with your king, and keep both pieces within reach of both wings so you can’t be stretched by a second front. If you set the blockade up early, you can then make waiting moves indefinitely.