Kakuro solving strategies

Forced combinations

Some entries have only one valid set of digits. A 2-cell entry with sum 3 must contain 1 and 2. Sum 4 is always 1 and 3. Sum 16 is 7 and 9. Sum 17 is 8 and 9. At the 3-cell level: sum 6 is 1+2+3, sum 7 is 1+2+4, sum 23 is 6+8+9, sum 24 is 7+8+9.

Memorize these. They appear in nearly every puzzle, and recognizing them saves you from working through combination lists manually. You still need cross-entry logic to figure out which digit goes in which cell, but knowing the digits gives you a head start.

Start with short entries

Two-cell entries have the fewest possible combinations. A 2-cell entry can have at most 4 different digit pairs for any given sum. Compare that to a 5-cell entry, which might have 10 or more valid combinations. Short entries are easier to reason about, and solving them feeds useful constraints to longer entries nearby.

Cross-entry elimination

Every white cell belongs to exactly one across entry and one down entry. The valid digits for that cell are the intersection of what both entries allow. If the across entry restricts the cell to {1, 3, 5} and the down entry restricts it to {3, 7}, the cell is 3.

This is where kakuro differs from solving entries in isolation. The power comes from the intersections. Even when each entry individually has several valid combinations, the crossing is often enough to pin down a cell.

Combination filtering

For each entry, list every digit combination that sums to the clue without repeats. Then eliminate combinations where any digit conflicts with a cell already filled in a crossing entry. If only one combination survives, all its digits are determined for that entry.

Even when multiple combinations remain, check which digits appear in every surviving combination. If every valid combination includes 5, then 5 must be in that entry. If only one cell in the entry can hold 5 (because the other cells have crossing constraints that exclude it), you have a placement.

The combinations chart is useful here. Having it open while solving saves time, especially for 4-cell and 5-cell entries.

Naked singles

When a cell has only one remaining candidate after all elimination, place that digit. This sounds obvious, but the power is in the chain reaction. Filling one cell reduces candidates in both its across and down entries. Those reductions may create nakeds singles elsewhere, which in turn trigger more eliminations. A single placement can cascade across the grid.

Use notes mode to track candidates. When a cell shows one digit, fill it immediately. Then re-scan the entries that cell belongs to for new singles.

Required values

If a specific digit appears in every valid combination for an entry, that digit must be placed somewhere in the entry. Check each cell to see which ones can hold it. If only one cell admits that digit (because the crossing entry for the other cells excludes it), you have a guaranteed placement without knowing the full combination.

This is especially useful for longer entries (4+ cells) where many combinations survive. Required values narrow things faster than trying to eliminate entire combinations.

Multi-entry cross-referencing

Basic cross-entry elimination looks at one across and one down entry for a single cell. Multi-entry cross-referencing extends this: if placing digit X in cell A (which is in entry E1) would make it impossible for entry E2 (which shares another cell with E1) to reach its target, then X is eliminated from cell A.

This is chain reasoning. You are looking two or three steps ahead: "If I put 4 here, the remaining cells in this entry would need to sum to 11 in 2 cells with an 8 already used, which is impossible." It takes practice, but it unlocks positions where simple elimination stalls.

Combination set intersection

When two entries share a cell and both have limited valid combinations, the shared cell must hold a digit that appears in at least one valid combination from each entry. List the possible digits for that cell from the across entry's combination set and from the down entry's combination set. The intersection is your candidate list.

For example, if the across entry's surviving combinations place {2, 5, 7} in the shared cell and the down entry's surviving combinations place {3, 5} in the same cell, the cell must be 5. This is more precise than basic cross-entry elimination because it accounts for the full constraint set of each entry, not just individual digit legality.

Systematic candidate tracking

On large grids, you cannot hold all constraints in your head. Use notes mode to write down every candidate for every empty cell. Work through entries methodically: update candidates after each placement, apply elimination across crossing entries, and scan for naked singles.

The approach is mechanical, but it works. Think of it as maintaining a spreadsheet of possibilities. Each time you eliminate a candidate or place a digit, ripple the update through all affected entries. Expert and einstein puzzles reward disciplined tracking over intuition.