Klondike Solitaire Rules — Draw One, Draw Three, and the Win Rate Nobody Agrees On

How Klondike Solitaire actually works: the seven columns, the two draw modes, why an empty column only takes a king, and what "winnable" really means for a deal.

Klondike is the solitaire game everyone means when they say "patience" or just "solitaire." It shipped with Windows for decades, which is why its rules feel like common knowledge even to people who have never picked up a deck. But the details — what an empty column is for, why draw three is so much harder than draw one, and how often a deal can even be won — are where most players guess wrong. Here is the game as it is actually dealt and played.

The board you are dealt

A single standard 52-card deck is all the game uses. At the start, 28 cards go into the tableau: seven columns, each one card longer than the last. The first column has one card, the second has two, and so on to the seventh with seven. In every column except the first, only the top card is face up; the rest sit face down until you uncover them. The remaining 24 cards form the stock, which you turn through one move at a time.

Four empty foundation piles sit above the tableau, one per suit. Those are the goal. Everything you do is in service of moving all 52 cards, suit by suit and in order, onto those four piles.

The three places every card can go

Every card you touch moves to one of three destinations:

• Another column. You build down in alternating colours — a red seven onto a black eight, a black six onto a red seven. The card you move, and any stack beneath it, must follow this rule.
• A foundation. You build up within a single suit, starting with the ace and ending with the king. Once a card is on a foundation it does not come back.
• The waste. When you turn a card from the stock and cannot or will not play it yet, it lands on the waste pile, where only its top card is available.

Almost every decision in Klondike is a choice about which of those three to do first, because the order you expose cards decides how many options stay open.

Draw one versus draw three

The stock can be turned in two ways, and the difference is enormous.

Draw one turns a single card from the stock to the waste. You can play that card immediately, and if you do not, it waits on top of the waste for your next move. This is the forgiving version — the card you need will come around again on the very next pass.

Draw three turns three cards at once. You may only play the top of those three; the other two are buried until the pile cycles all the way around. Because a card can sit two layers deep in the waste for an entire pass, draw three is markedly harder — estimates put the win rate well below the draw-one game, sometimes by half.

Both modes deal from the same 52 cards, so the choice is purely about difficulty. If you are learning, start on draw one; move to draw three once the game stops feeling like a puzzle you can always solve.

Why an empty column only takes a king

An empty column is the most valuable square on the board, and the rules protect it. You may place a card on an empty column only if that card is a king. Nothing else — not a queen, not a pile of cards ending in a king — may start a new column.

This is deliberate. A king begins a fresh descending sequence, and because a king cannot be placed on anything else, an empty column is the only place a king-led pile can live. Emptying a column is therefore a resource: do it only when you have a king (or a king-led sequence) ready to drop in. Emptying a column just to expose one card, with no king to fill it, often costs you more than it gains.

How many deals can you actually win

This is the question with no clean answer, and anyone who quotes a single number is guessing. Klondike deals are not all winnable — unlike, say, a properly generated Sudoku, the initial layout can be unsolvable before you touch a card. A widely cited estimate is that roughly four deals in five can be finished with perfect play, and the remaining fifth cannot be won no matter what you do. Draw three cuts that figure substantially because the waste buries cards you cannot reach.

The honest version: a loss is usually not bad luck, but a win is never guaranteed. The deal sets a ceiling; your play determines how close you get to it. Undo is the tool that lets you find that ceiling — when a line stalls, step back and try a different sequence rather than resigning the board.

The habits that win more games

A few moves tilt the odds in your favour more than the rest:

• Expose hidden cards first. A face-down card is a question mark; every move that turns one up is progress even if it looks small.
• Hold empty columns for kings. Do not spend them early on a low card you could have placed elsewhere.
• Send aces and twos to the foundations early. They free up space in the tableau and rarely help you build downward.
• Turn the stock before rearranging piles you have already seen. New information beats re-sorting known cards.

None of these guarantees a win, but together they push most deals toward the winnable side of that four-in-five line.