MULTINOMIAL in Excel: Count Arrangements of Repeated Categories

• Introduced: First introduction version is not established by the official history reviewed.
• Availability: official help lists Excel for Microsoft 365, Excel for Microsoft 365 for Mac, Excel 2024, Excel 2024 for Mac, Excel 2021, Excel 2021 for Mac, Excel 2019, Excel 2016.
• Checked: October 14, 2026; official MULTINOMIAL reference. Documentation and independently calculated examples; no native Excel execution or broader platform certification.

=MULTINOMIAL(number1,[number2],...)

Microsoft documents one required argument and up to 255 supplied values in total. Negative arguments produce #NUM! and nonnumeric arguments produce #VALUE! according to the dedicated reference.

The examples use explicit nonnegative integer category counts. The current page does not establish a complete fractional-input truncation or range-text matrix, so neither is inferred from FACT or another spreadsheet’s similarly named function. Optional brackets belong to the explanatory syntax, not the entered formula.

For A, A and B:

=MULTINOMIAL(2,1)

Expected count: 3, from AAB, ABA and BAA. Swapping the two A copies is invisible under this model and does not add new outcomes. Treating all three objects as distinct would give FACT(3)=6 instead.

For A, A, B and C, the count becomes 12. The two identical A labels create a factor-of-two overcount in the 24 permutations of four separately identified objects.

For category counts 3, 2 and 1:

=MULTINOMIAL(3,2,1)
=FACT(3+2+1)/(FACT(3)*FACT(2)*FACT(1))

Both expected results are 60, from 720/(6×2×1). The arguments are counts within categories, not category labels or probabilities. Their sum gives the total number of positions.

A useful small table is counts (1,1), (2,1), (2,2), (3,2,1), with expected results 2, 3, 6 and 60. Keep each tuple as a separate formula so argument boundaries are unambiguous.

For three A labels, two B labels and one C label, choose the three A positions among six. Then choose the two B positions among the three remaining slots:

=COMBIN(6,3)*COMBIN(3,2)

Expected result: 60, from 20×3. The final slot is forced to C. This independent construction agrees with the multinomial count and shows why the category counts belong to the same total set of positions.

Choosing both categories independently from all six slots would allow overlapping assignments and overcount invalid arrangements. Update the remaining capacity at each stage.

For counts 2, 1 and 1, the result is 12. If the two single-count categories become indistinguishable for the report, merge them into a count of two:

=MULTINOMIAL(2,2)

Expected result: 6. The reduced count reflects a changed definition of a distinguishable outcome, not a numerical rounding effect.

Adding an explicitly zero-size category to the integer tuple 2,1,1 leaves the mathematical coefficient at 12 because FACT(0)=1. That identity is an independently derived boundary expectation; no new native result is claimed for every omitted, empty-cell or zero-argument configuration.

Confirm that all supplied counts refer to disjoint categories and add to the intended number of positions. A result alone cannot identify a misclassified object or decide whether two labels should be treated as identical.

The factorial ratio is useful for small verification cases, but large intermediates grow rapidly. Do not use the arithmetic model as a guarantee of native overflow limits or exact representation for every large arrangement count. Keep all count and type assumptions visible beside the formula.

• Microsoft MULTINOMIAL reference.
• Editorial testing standards.

Worked inputs are fictional. Decimal answers are independently calculated mathematical expectations rounded to nine places, not captured native results. Extreme magnitudes, complete coercion matrices and cross-version numerical accuracy remain untested.