How this instrument works
A percentile is a value below which a given percentage of your sorted data falls. The 90th percentile is the value with about 90% of the data at or below it; the 40th percentile has about 40% below it. Percentiles generalise quartiles — Q1, the median and Q3 are simply the percentiles at 25, 50 and 75 — but this instrument lets you ask for any of the other 97 cut points too, which is what most real percentile questions actually need.
This calculator finds the value at your chosen percentile by linear interpolation between the two closest ranked values, the method NIST documents and spreadsheets use in their PERCENTILE.INC function: sort the data, compute the rank position 1 + (p/100) x (n - 1), and interpolate between the neighbouring values whenever that position falls between two ranks. It is one of several published percentile conventions; this one is the common spreadsheet default, and being explicit about it matters because different methods can shift small-sample answers by a noticeable amount.
Percentiles turn up anywhere a raw number needs context: a race finish time only means something once you know it beat 80% of the field; a test score matters alongside where it lands among everyone else who took it; a response-time monitoring dashboard flags its 95th or 99th percentile because a handful of slow outliers can hide in an average. This instrument answers the direction most people actually need day to day — given a percentile, what's the value — rather than the reverse question of where a given value ranks.
- Paste or type your numbers into the Data set field — commas, spaces or new lines all work.
- Set Percentile (0-100) to the cut point you want — it opens at 90, a common tail-risk or top-performer threshold.
- Read Value at the pth percentile beneath it — it recalculates the instant either field changes.
- Want the standard quarter marks instead of a custom percentile? This site's quartile calculator returns Q1, the median and Q3 in one step using the same method.
Worked example — the 90th percentile of eight quiz scores
Take {2, 4, 4, 4, 5, 5, 7, 9}, already sorted, n = 8, with Percentile set to 90. The rank position is r = 1 + (90/100) x (8 - 1) = 1 + 6.3 = 7.3, which falls between the 7th value (7) and the 8th value (9) in the sorted list.
Interpolating 0.3 of the way between them: value = 7 + 0.3 x (9 - 7) = 7 + 0.6 = 7.6. So 90% of these eight quiz scores fall at or below 7.6 — only the top score, 9, sits above it.
Questions
What's the difference between this and a percentile rank calculator?
This instrument runs in one direction: you give it a percentile (say, 90) and it returns the matching value from your data. A percentile rank calculator runs the reverse — you give it a value and it tells you what percentage of the data falls below it. The two are related but answer different questions.
How is a percentile different from a quartile?
A quartile is just one of three specific percentiles — the 25th, 50th and 75th. Percentile is the general term covering all 99 cut points from 1 to 99 (or any point in between, including decimals). If you specifically want Q1, the median and Q3, this site's quartile calculator is the more direct tool; use this page when you need a percentile that isn't one of those three.
Why did I get a value that isn't in my original data set?
Because the method interpolates. Unless your chosen percentile's rank position lands exactly on one of your sorted values, the instrument reports a point between two neighbouring values, weighted by how close the rank sits to each — which is normal and matches how spreadsheet PERCENTILE functions behave, not a computation error.
Can I enter a percentile like 99.5 or 12.5?
Yes — the Percentile field accepts any value from 0 to 100, including decimals, in steps of 0.1. That flexibility is the main reason to use this tool over the fixed quartile calculators, which only report the three quarter marks.
Why do percentile results vary between calculators for the same data?
There are several published interpolation conventions for percentiles, and different tools default to different ones. This instrument uses linear interpolation between closest ranks — rank = 1 + (p/100)(n-1) — the same method as Excel and Google Sheets' PERCENTILE.INC function and the one NIST documents, so results should match any tool using that same convention.
What does the 0th or 100th percentile give me?
The 0th percentile returns your data set's minimum value, and the 100th percentile returns its maximum — both are the interpolation formula's endpoints, since the rank position collapses to the very first or very last ranked value with no interpolation needed.