What Does an IQ of 90 Mean? Is It Low, Average or Below Average?

An IQ of 90 is the 25th percentile on the standard scale (mean 100, standard deviation 15): about 1 person in 4 scores 90 or lower and 3 in 4 score higher. It is not a low score. The Wechsler children's test (WISC-V), the Stanford-Binet 5 and the Woodcock-Johnson all put 90 at the bottom of their Average band, and half of all people score between 90 and 110. It is “below average” only in the literal sense of sitting 10 points under 100, and some simpler charts label the 85 to 99 stretch “low average”. This guide shows where 90 sits on the bell curve, how the labels differ by chart and by scale, how far to trust one score and what a score of 90 does and does not predict.

• Percentile: 25th (25.2%) on a standard deviation 15 scale; 26.6th on standard deviation 16 and 33.8th on standard deviation 24 (our normal-curve arithmetic).
• Rarity: not rare. About 1 person in 4 scores 90 or lower, which is 7 or 8 pupils in a class of 30.
• Label: Average on the WISC-V and Stanford-Binet 5 (90 to 109) and on the Woodcock-Johnson (90 to 110); low average on simpler charts that call 85 to 99 low.
• Accuracy: read a reported 90 as roughly 85 to 95.
• Prediction: a meta-analysis found IQ correlates .56 with education, .43 with occupation and .20 with income (Strenze, 2007): tendencies in groups, not limits on a person.

The figure shows why 90 matters on the bell curve: the quartile lines fall at 89.9 and 110.1, so 90 is the point where the middle half of the population begins. Our guides to an IQ of 85 and an IQ of 115 cover the scores either side.

What percentile is an IQ of 90?

On the standard scale a 90 sits two thirds of a standard deviation below the average (z = -0.67), which puts it at the 25th percentile. The table places the scores around it on the same curve, and our IQ percentile calculator does the same for any score.

The printed number depends on the scale. Most modern tests use a standard deviation of 15, but older Stanford-Binet editions used 16 and the Cattell scales use 24, and the same performance prints as a different number on each. Our explainer on standard deviation 15 versus 16 walks through the arithmetic, and the IQ score converter moves a score between scales. A percentile is a position within a norm sample of people, which is also why animals have no IQ score; our guide to elephant IQ shows what researchers measure instead.

Is an IQ of 90 low, average or below average?

It depends on the chart, so check which chart a label comes from. The table shows where 90 falls on the charts we cover.

Three of the four charts call 90 plainly average, and the fourth splits its average range at 100. None of them treats 90 as a concern in itself. The detailed band tables are in our guides to WISC-V scores, the Stanford-Binet and the Woodcock-Johnson, and the simplified ranges are in IQ ranges and IQ classifications.

“Below average” is literally true, because 90 is under 100, and so is the score of half the population. It is also 20 points above the reference point of about 70 used in assessing intellectual disability, a diagnosis that needs evidence of difficulty in everyday functioning as well as a score; our guide to an IQ of 70 explains it.

How accurate is a score of 90?

Not very, at the level of single points. Every score carries a standard error of measurement, commonly 2 to 3 points on a full-scale adult battery, which gives a 95% interval of about 5 points either side. A reported 90 is best read as 85 to 95. That range crosses the line between the 80 to 89 and 90 to 109 labels, so whether a report prints “Low Average” or “Average” for someone near 90 can come down to chance on the day. Our explainer an IQ score is a range, not a number shows how to read the band, and how to read an IQ test report shows where a real report prints it. A score from an online test depends on how that test was normed, so compare it with a clinical report only with care.

Does the age of the test change a score of 90?

It can. Test norms go out of date because raw performance has risen over the last century. A meta-analysis of 271 samples from 1909 to 2013 estimated gains of 0.28 full-scale IQ points a year, roughly 3 points a decade, and found that the gains have decreased in more recent decades (Pietschnig and Voracek, 2015). As an illustration only, twenty years at 0.28 a year is 5.6 points, which would turn a 90 on stale norms into roughly an 84 on fresh ones. The real gap depends on the country and the test, and the effect has slowed, so treat the arithmetic as a reason to ask which edition produced a score. Our news piece on why IQ norms expire covers the evidence.

What happens when a score of 90 is retested?

Scores regress toward the average when they are repeated, because part of what put the first score below 100 was chance. For two tests on one scale, the expected second score is 100 plus the correlation between the tests times the distance of the first score from 100, which for a 90 is minus 10. The table applies that formula using correlations reported in studies of how stable cognitive scores are.

These are group averages on an idealised model, not predictions for one person, and an individual's score can move in either direction. The first three rows come from one study of 11,119 children whose tests changed with age (what 11,119 children show), so a change in the test itself adds to the shift. The adult row is the mean rank-order stability for age 20 and a five-year gap in a meta-analysis of 205 longitudinal studies (Breit and colleagues, 2024). Our page on regression to the mean explains the effect.

What does an IQ of 90 predict?

Test scores predict later outcomes in groups, but loosely. A meta-analysis of longitudinal studies found correlations of .56 with educational attainment, .43 with occupational level and .20 with income (Strenze, 2007). Squared, those figures say how much of the spread in an outcome goes with the score: about 31% for education, 18% for occupation and 4% for income. Most of the variation in how people do sits outside the test, so a 90 nudges the odds and sets no ceiling.

For the detail behind those numbers, see does IQ predict school success, does IQ predict income and average IQ by education level. For how a college entrance score maps onto this scale, see SAT score to IQ, and for the average itself see what the average IQ test score is.

Quick answers about an IQ of 90

• Is an IQ of 90 low? Not on the major charts. It is the 25th percentile and the first score in the Average band on the WISC-V, the Stanford-Binet 5 and the Woodcock-Johnson.
• Is 90 below average? Literally yes, because it is under 100, and so is the score of half the population.
• How rare is an IQ of 90? It is not rare: about 1 person in 4 scores 90 or lower.
• What percentile is 90? The 25th on a standard deviation 15 scale, the 26.6th on standard deviation 16 and the 33.8th on standard deviation 24.
• Is 90 an intellectual disability? No. The reference point is about 70, and a diagnosis also needs evidence of difficulty in everyday functioning.
• How accurate is it? Read a reported 90 as roughly 85 to 95.

Sources

• Strenze, T. (2007): Intelligence and socioeconomic success: a meta-analytic review of longitudinal research, Intelligence, 35(5), 401-426.
• Pietschnig, J. and Voracek, M. (2015): One century of global IQ gains: a formal meta-analysis of the Flynn effect (1909-2013), Perspectives on Psychological Science, 10(3), 282-306.
• Breit, M., Scherrer, V., Tucker-Drob, E.M. and Preckel, F. (2024): The stability of cognitive abilities: a meta-analytic review of longitudinal studies, Psychological Bulletin, 150(4), 399-439.
• Percentile and rarity figures are computed from the normal distribution. The bands, the scale conversion and the standard error of measurement are those described in our own guides to the Stanford-Binet, the Woodcock-Johnson, WISC-V scores and IQ score margin of error.