Account

What new tool do you want?

Math & Education

Percentile Calculator

Find the value at any percentile of a data set, or the percentile rank of a value — using the same method as Excel's PERCENTILE.INC.

How to use

  1. 1Paste your numbers.
  2. 2Enter a percentile (e.g. 90) to find its value, or a value to find its percentile rank.

How it's calculated

Value at percentile p (inclusive, like Excel PERCENTILE.INC): rank = p × (n − 1) on the sorted data, interpolating between neighbours. Percentile rank of x = share of values below x plus half of those equal.

Frequently asked questions

Why do different tools give different percentiles?

There are several definitions (inclusive, exclusive, nearest rank). This one matches Excel's PERCENTILE.INC and Google Sheets' PERCENTILE.

What are quartiles?

The 25th (Q1), 50th (median) and 75th (Q3) percentiles. The interquartile range Q3 − Q1 covers the middle half of the data.

Is the 90th percentile 90% of the maximum?

No — it's the value below which about 90% of the data falls.

Interpolation is why percentiles aren't always one of your actual values

For the data set 15, 20, 35, 40, 50, the 40th percentile comes out to 29 — a number that isn't in the list at all. The method (matching Excel's PERCENTILE.INC) places the percentile at a fractional rank in the sorted data — rank = p × (n − 1), so 0.4 × 4 = 1.6 — then interpolates 60% of the way between the 2nd value (20) and the 3rd (35), landing at 20 + 0.6 × 15 = 29.

Percentile rank handles ties by splitting the difference

Going the other way — finding what percentile a given value sits at — counts how many values are strictly below it, plus half credit for any values exactly equal to it, divided by the sample size. For the set 1, 2, 3, 4 and value 3, one value (1, 2) sits below and one value (3 itself) ties, giving (2 + 0.5) ÷ 4 = 62.5% rather than a plain 75% or 50% — the half-credit for ties keeps the measure symmetric.

Why another tool might give you a different number

PERCENTILE.INC is only one of several accepted definitions — PERCENTILE.EXC, nearest-rank, and various textbook methods all compute slightly different values for the same data, especially in small samples. None is 'more correct'; they're different conventions, so when comparing results across tools it's worth checking which definition each one uses, particularly for quartiles (Q1, median, Q3) where the discrepancy is most visible.

A blank rank box leaves out the rank, and a stray word stops everything

The value-to-rank field starts empty, and while it is empty I only see the percentile I asked for plus the lower quartile, the median, and the upper quartile. Typing a number adds one more line, the rank of that number, and clearing it removes the line again instead of ranking zero. The data box accepts commas, spaces, or semicolons between values, and the list is sorted before any percentile is taken, so the order I pasted does not matter. A word or a double comma that leaves a token which is not a number stops the result and quotes that token back to me. The percentile itself must sit from 0 through 100 inclusive.

Related tools

More: Math & Education