Percentile vs Percentage: What the Difference Really Means
Scoring 90% on a test and being in the 90th percentile sound like the same achievement. They are not — and in a hard exam they can be wildly different numbers. One describes your performance against the paper; the other describes your position against everyone else who sat it.
A percentage is a share of a total. A percentile is a rank within a group. Scoring 90% means you got 90 of every 100 marks. Being in the 90th percentile means you scored higher than 90% of the other people. On a difficult exam, 62% could easily be the 90th percentile.
The core difference
| Percentage | Percentile | |
|---|---|---|
| Measures | Amount out of a total | Position in a ranked group |
| Depends on | Only your own result | Everyone else's results |
| Range | Can exceed 100% | Always 0–100 |
| Example | "You answered 45 of 60 = 75%" | "You beat 82% of candidates" |
The practical consequence: a percentage is stable, a percentile moves. Take the same paper twice with a stronger cohort and your percentage stays put while your percentile falls.
How to calculate a percentile
The standard definition counts how many values fall below yours:
Percentile = (Number of values below yours ÷ Total number of values) × 100
Some sources use a variant that gives half credit for ties: (below + 0.5 × equal) ÷ total × 100. This matters in small groups with repeated scores, and it is the version most statistics packages use by default.
A worked class example
Twenty students sit a test. You score 68, and 15 students scored below you, 2 tied with you, and 2 scored higher.
- Simple method: 15 ÷ 20 × 100 = 75th percentile.
- With ties: (15 + 1) ÷ 20 × 100 = 80th percentile.
Your percentage score, meanwhile, depends entirely on the marks available. If the paper was out of 80, you scored 68 ÷ 80 = 85%. Two different numbers, both correct, describing two different things.
The percentage is fixed by the marks available. The percentiles move with the cohort and with the counting method you choose.
Where percentiles are used
- Standardised tests. SAT, GRE and IQ reports lead with percentiles because raw marks are not comparable across different paper versions.
- Child growth charts. A baby on the 40th percentile for weight is heavier than 40% of babies the same age — it is not a score, and it is not a target.
- Salary benchmarking. The 75th percentile salary for a role is the figure that a quarter of people in that role exceed.
- Performance monitoring. Engineers quote p95 and p99 response times because averages hide the slow tail.
Common misreadings
- Treating the 50th percentile as "average". It is the median, which only equals the mean in a symmetric distribution.
- Reading a low growth percentile as a problem. Half of all healthy children are below the 50th by definition.
- Averaging percentiles. Ranks cannot be meaningfully averaged; go back to the underlying values first.
- Confusing percentile with percentage points. Moving from the 60th to the 70th percentile is a 10 percentile-point move — see percent vs percentage points.
Practice problems
1. In a group of 50, your score beats 38 people. What percentile are you in?
2. You scored 34 out of 40. What is your percentage?
3. A salary sits at the 90th percentile of a 200-person band. How many earn more?
4. Can you be in the 100th percentile?
Frequently Asked Questions
Found this useful?
Share it with someone who should start investing today.