The IQ scale explained: what 85, 100, 115 and 130 mean

An IQ score only makes sense once you know its scale. Almost every modern test uses the same one: the average is set to 100 and the standard deviation (SD) to 15. Scores are then spread out along a bell curve.

The bell curve in one picture

If you lined up a large, representative group of people by their test performance, most would cluster near the middle and fewer would sit at either end. On the IQ scale:

• about 68% of people score between 85 and 115 (within one SD of the average),
• about 95% score between 70 and 130 (within two SDs),
• about 2.3% score above 130, and about 2.3% below 70.

What each score means

The table shows the share of the reference group that scores below each value. That share is the percentile.

So a score of 115 does not mean “15% smarter than average”. It means the person did better than roughly 84 out of 100 people in the comparison group. Try any number in the IQ percentile calculator.

Where the number comes from

Early tests used a ratio IQ: mental age divided by real age, times 100. A 10-year-old performing like a typical 12-year-old scored 120. That works poorly for adults, because mental age stops growing.

From 1939 David Wechsler popularised the deviation IQ, which is still used today. Your raw score is compared with people of your age, then converted to a scale with mean 100 and SD 15. That is why every IQ score is a statement about rank, not about an absolute quantity.

Not every test uses SD 15

Most tests do, but not all. Some older tests used an SD of 16, and some report scores with an SD of 24. On an SD-24 scale, the same top-2% rank becomes 148 instead of 130. When you compare scores from different sources, check the scale first. This site always reports on the standard SD-15 scale.

Labels vary, so be careful with them

Test publishers attach words such as “average” or “high average” to score bands, and the cut-offs differ from one test to another. Words like “genius” are not scientific categories at all. A percentile is easier to understand and more honest.

Why a range is more honest than a single number

Every test has measurement error. On a long clinical battery, the error band around a score is a few points. On a short test it is much wider: a 16-item test can only place someone to within about ±13 points, and a 32-item test to about ±10.

That is why the result on this site is an estimated range, such as “105 to 131”, rather than a falsely precise “118”. The methodology page shows the exact calculation.

The ends of the scale

Short tests cannot rank the extremes well, because very few items are hard or easy enough to tell people apart there. If your range reaches the top or bottom of what this test can measure, the result says so (“135 or higher”) instead of inventing a number.

Curious where you land? Take the free quick test and see your estimated range instantly, with no email and no payment.