Percentile to Z Score Calculator
Run the normal distribution backwards: enter a percentile and get the matching z-score, tail probabilities, and the shaded curve. Free inverse z-score calculator — no signup.
Convert percentile to z-score
Enter the percentile rank — e.g. 95 for the 95th percentile, 2.5 for the 2.5th.
The shaded area matches your percentile; the red marker is the resulting z-score.
This percentile to z score calculator runs entirely in your browser. Nothing is uploaded or stored.
What "inverse" means here
The usual z-score workflow goes forward: raw score → z-score → percentile. The inverse z score calculator works backwards from an area to the z-score that produces it. You hand it a percentile — say, the 95th — and it finds the point on the bell curve with exactly 95% of the area to its left. That point is z = 1.645.
Mathematically this is the inverse of the normal cumulative distribution function (sometimes called the standard normal quantile function). There is no pencil-and-paper formula for it — historically you read it from the z-table in reverse, scanning the body of the table for the probability nearest your target and reading the z-score off the margins. The converter above performs that inverse calculation directly, to four decimal places, and draws the matching shaded region so you can see the area-to-score relationship.
Note the direction carefully: our z score to percentile calculator answers "this z-score is which percentile?", while this page answers "this percentile is which z-score?" They are mirror images, and using the wrong one is a classic exam error.
Worked examples
Example 1 — the 95th percentile. Divide by 100: 0.95. Find the z-score with a left-tail probability of 0.95 → z = 1.645. So the cutoff for the top 5% of a normal distribution is 1.645 standard deviations above the mean. If a graduate program admits the top 5% of test-takers and scores have mean 500 and SD 100, the cutoff is 500 + 1.645 × 100 = 664.5.
Example 2 — the 2.5th percentile. 2.5 / 100 = 0.025. The z-score with left-tail probability 0.025 is z = −1.96. This is the lower boundary of a 95% confidence interval, mirrored by +1.96 on the upper side.
Example 3 — the 50th percentile. 0.50 → z = 0.0000, exactly the mean. If your inverse calculation does not return zero here, check that you entered the percentile (50) and not the probability (0.5) — the converter accepts the 0–100 scale.
Critical values reference
The percentiles below are the ones statistics courses test most. Keep this z critical value table bookmarked — or just use this page as a z critical value calculator whenever you need one:
| Percentile | Z-score | Common use |
|---|---|---|
| 90th | 1.282 | One-tailed 10% cutoff |
| 95th | 1.645 | One-tailed 5% cutoff |
| 97.5th | 1.960 | Two-tailed 5% cutoff · 95% CI |
| 99th | 2.326 | One-tailed 1% cutoff |
| 99.5th | 2.576 | Two-tailed 1% cutoff · 99% CI |
Notice the pair students mix up most: the 95th percentile is z = 1.645, while the 97.5th percentile is z = 1.96. The 95% confidence interval uses 1.96 because it needs 2.5% in each tail — the middle 95%, not the bottom 95%.
When you need the inverse conversion
- Finding cutoff scores. "Top 10% get an A" → the 90th percentile → z = 1.282 → cutoff = mean + 1.282 × SD. Any grading curve or admission threshold stated as a percentile starts here.
- Building confidence intervals. A 95% interval needs the z-score that leaves 2.5% in each tail: convert 97.5 → 1.96. A 99% interval needs 99.5 → 2.576.
- Quality-control limits. "Flag the most extreme 1% of parts" → convert 99.5 → ±2.576 standard deviations from the target as the control limits.
- Power and sample-size planning. Researchers convert target percentiles to z-scores before plugging them into sample-size formulas.
- Checking published claims. When a report says "top 5%", convert it to z = 1.645 and see whether the underlying numbers support the claim.
The by-hand method: reading a z-table backwards
Before inverse calculators existed, you did this conversion with a printed z score table — and the technique is still worth knowing, because exams love it. The process runs the normal table lookup in reverse:
- Convert your percentile to a probability. The 95th percentile becomes 0.9500.
- Scan the body of the table for the value closest to 0.9500. You will find 0.9505 sitting in the table.
- Read the z-score off the margins: row 1.6, column .04 or .05 — about z = 1.645.
The weakness of the by-hand method is precision: tables round to four decimals and the exact value you want rarely appears, so you interpolate between neighbors. The converter above skips all of that — it computes the inverse normal function directly, so z = 1.6449 comes out exact rather than "somewhere between 1.64 and 1.65". Learn the table method once for the concept; use the calculator for real answers.
One-tailed vs two-tailed cutoffs
The most expensive confusion in inverse problems is mixing up one-tailed and two-tailed cutoffs. A question that says "the top 5%" is one-tailed: convert the 95th percentile → z = 1.645. A question that says "the middle 95%" is two-tailed: you need 2.5% in each tail, so convert the 97.5th percentile → z = ±1.96. Same "95%", two different z-scores — and only the wording tells you which one applies. When a question says "95% confidence interval", it means the middle 95%, so reach for 1.96. When it says "95th percentile", it means the bottom 95%, so reach for 1.645. The converter's shaded curve shows you which interpretation you just computed, which makes the distinction visible instead of abstract.
Frequently asked questions
How do you convert a percentile to a z-score?
Work backwards from the area: find the z-score whose left-tail probability equals the percentile divided by 100. The 95th percentile gives z = 1.645; the 97.5th gives z = 1.96. The calculator above performs the inverse normal calculation instantly and shades the matching area on the curve.
What z-score is the 95th percentile?
The 95th percentile corresponds to z = 1.645 — meaning 95% of values fall at or below it and 5% above. Note this differs from 1.96, which marks the 97.5th percentile and is the two-tailed 95% cutoff. Mixing these two up is one of the most common statistics mistakes.
What is an inverse z-score calculator?
It reverses the usual direction: instead of entering a z-score to get a percentile, you enter a percentile (or tail probability) and get the z-score back. It is the tool you need when a question gives you "the top 5%" and asks for the cutoff score.
What z-score gives a 99% confidence interval?
For a two-tailed 99% confidence interval, the critical z-score is 2.576 — leaving 0.5% in each tail. For a one-tailed 99% cutoff (top 1%), it is 2.326. The critical-values table on this page lists the standard cutoffs so you never have to memorize them.
Why would I need a percentile-to-z conversion?
Whenever a threshold is stated as a percentile: admission cutoffs ("top 10%"), grading curves, quality-control limits, and confidence-interval construction all start from a percentile and need the matching z-score before any further calculation can proceed.