IQ scale: what each score range means
On the scale most modern tests use, the average is 100 and the standard deviation is 15. Scores from 90 to 109 are called average and cover about half of all people. Higher and lower scores are grouped into bands of about ten points, each with a descriptive label. The bands are a convenient way of talking about results, not sharp borders between kinds of people.
The usual score bands
The table shows the bands commonly used on a deviation-15 scale, with the approximate share of people in each. The shares come from the normal curve and are rounded.
| Score | Common label | Approximate share of people |
|---|---|---|
| 130 and above | Very high | about 2% |
| 120 to 129 | High | about 7% |
| 110 to 119 | High average | about 16% |
| 90 to 109 | Average | about 50% |
| 80 to 89 | Low average | about 16% |
| 70 to 79 | Low | about 7% |
| Below 70 | Very low | about 2% |
The scale is symmetrical. The band from 110 to 119 holds about as many people as the band from 80 to 89, and the two outer bands are equally rare. If you want the exact position of a single score rather than a band, see the IQ percentile table.
What the bands mean in practice
- 90 to 109. The middle half of the population. Most people you meet score here.
- 110 to 119 and 80 to 89. Clearly above or below the middle, but still common: about one person in six is in each of these bands.
- 120 to 129 and 70 to 79. Less common. Roughly one person in fifteen is in each.
- 130 and above, and below 70. The two ends of the curve, each with a little over 2% of people. A score of 130 is the usual threshold for Mensa, provided it comes from an approved, supervised test.
A low score on its own is not a diagnosis. Clinical decisions are made by psychologists who look at much more than one number, including how a person manages in daily life.
Labels differ between publishers
The numbers in the left column are fairly standard. The words in the middle column are not. Each test publisher chooses its own labels, and they have changed over time. Older manuals described the top band as very superior and the band from 70 to 79 as borderline. Newer editions tend to use plainer words such as extremely high, very high, very low and extremely low.
Some historical labels for low scores are now considered offensive and have been dropped altogether. The word genius, which appears on many charts on the internet, is not an official band on any major clinical test.
So if two reports use different words for the same score, nothing about the result has changed. Look at the number and the scale it is on, not at the adjective.
Different scales give different numbers
A score only makes sense together with its standard deviation. Most modern tests, including the Wechsler scales, use 15. Older editions of the Stanford-Binet used 16, and the Cattell scale uses 24. All of them put the average at 100, but they stretch the distance from the average differently.
The result is that the same position among other people is written as a different number on each scale:
| Position | Deviation 15 | Deviation 16 | Deviation 24 (Cattell) |
|---|---|---|---|
| About the 2nd percentile | 70 | 68 | 52 |
| About the 16th percentile | 85 | 84 | 76 |
| 50th percentile | 100 | 100 | 100 |
| About the 84th percentile | 115 | 116 | 124 |
| About the 98th percentile | 130 | 132 | 148 |
| About the 99.9th percentile | 145 | 148 | 172 |
A score of 148 on the Cattell scale and a score of 130 on a deviation-15 scale describe the same thing: a result higher than about 98% of people. This explains many of the very large numbers quoted in casual conversation. Before comparing two scores, check which scale each one is on. If you do not know the scale, the number alone tells you little.
A score is a range, not an exact value
No test measures perfectly. The same person taking the same test twice will not get exactly the same result, because attention, mood, tiredness and luck with particular questions all vary. This is called measurement error, and every serious test reports it.
For that reason psychologists usually give a result as a range. Even on a well-made, individually administered test, a reported score of 112 is better read as somewhere around 107 to 117. Shorter and less controlled tests have wider ranges.
This matters most near the borders of the bands. A person who scores 109 and a person who scores 111 have different labels in the table, but there is no real difference between them. On another day their results could easily be the other way round.
An online test, ours included, gives an estimate with a wider margin than a clinical test. It is not a clinical assessment and cannot diagnose anything. See how accurate online IQ tests are.
How to read your own result
- Check the scale. If the standard deviation is not 15, convert the score or use the percentile instead.
- Think of a range of a few points on either side of the number.
- Read the band as a rough description, and ignore the exact wording of the label.
- Remember what the score leaves out. It says nothing about creativity, character or effort, as explained in what is IQ.
If you want to see roughly where you land, you can take our test. For a formal result, talk to a qualified psychologist.
Keep reading
IQ percentile: what it means and how to read it
An IQ percentile tells you what share of people score below you. A conversion table from score to percentile and rarity, and why it beats the raw number.
How do IQ tests work?
What tasks an IQ test contains, how a raw score becomes an IQ, what makes a test reliable and valid, and how clinical tests differ from online ones.
Are online IQ tests accurate?
A well-built online IQ test can estimate your reasoning ability, but it is not a clinical assessment. What limits accuracy and how to spot a fair test.