Short answer: An IQ of 138 is very superior: higher than about 99.4 percent of people, roughly 1 in 177. It sits 2.53 standard deviations above the mean of 100 and qualifies for Mensa on an approved, supervised test. It is among the strongest predictors psychology has of academic success, yet the most famous study of high scorers shows it cannot forecast who will change a field.
Shaded: the share of people who score higher on a mean-100, SD-15 scale.
At a glance
| Percentile | 99.44th |
|---|---|
| Rarity | About 1 in 177 (0.565 percent score 138 or higher) |
| Wechsler classification | Very Superior (WAIS-IV) / Extremely High (WAIS-5) |
| Stanford-Binet 5 classification | Gifted or very advanced |
| On an SD 16 scale | 141 |
| On an SD 24 (Cattell) scale | 161 |
| Mensa eligible? | Yes, eight points above the 130 line on SD 15 tests |
What an IQ of 138 means
A 138 lands 38 points above average, 2.53 standard deviations up the bell curve. In a stadium of 50,000 people, about 280 would be expected to score this high or higher.
The WAIS-IV calls it Very Superior; the WAIS-5 calls the same band Extremely High. The Stanford-Binet 5 labels it gifted or very advanced. All three put 138 inside the top tier of their scales. For every label from 40 upward, see our IQ classification guide.
How 138 compares with its neighbors
| IQ | Percentile | Share at or above | Rarity |
|---|---|---|---|
| 130 | 97.7 | 2.3% | 1 in 44 |
| 135 | 99.0 | 0.98% | 1 in 102 |
| 138 | 99.4 | 0.565% | 1 in 177 |
| 140 | 99.6 | 0.38% | 1 in 261 |
| 145 | 99.87 | 0.14% | 1 in 741 |
Between 130 and 138, the qualifying pool shrinks by three-quarters. That is the steep side of the curve at work. More comparisons are on our IQ rarity chart.
Terman's gifted children: a century of evidence
In 1921 Lewis Terman of Stanford began tracking California schoolchildren with very high Stanford-Binet scores. By 1928 the sample held 1,528 children, 856 boys and 672 girls. His main selection line was an IQ of 140; the group averaged about 151.
The children did well. Well over half finished college, when about 8 percent of the general population did. A few became well known, among them the psychologist Lee Cronbach and the television writer Jess Oppenheimer. Most led solid but unremarkable lives.
Two boys screened for the study were not selected because their scores fell short. William Shockley shared the 1956 Nobel Prize in Physics; Luis Alvarez won it in 1968. According to psychologist Dean Keith Simonton, no member of Terman's sample won a Nobel in any field.
What that result does and does not prove
It does not prove that IQ is irrelevant. Psychologist Russell Warne argues that missing two future laureates was statistically expected: Nobel Prizes are so rare that no childhood test could predict them, and a single cutoff guarantees misses near the line.
It does show the limit of a number like 138. Scores forecast averages across large groups well. Individual breakthroughs depend on field, training, persistence, timing and luck. A 138 is a strong starting position, not a prediction.
Is 138 a good IQ?
It is an excellent one. In a study of more than 70,000 English students, Ian Deary and colleagues found that general intelligence at age 11 correlated 0.81 with national exam results at 16. At the level of adult careers, Tarmo Strenze's 2007 meta-analysis found moderate links with education (.56) and occupation (.43), and a weaker one with income (.20).
Those correlations describe groups. A 138 raises the odds of academic success; it does not settle any individual outcome. For how this range fits the wider top tier, see the top 1 percent IQ guide.
How to verify a score of 138
A defensible 138 comes from an individually administered, normed test, such as the WAIS-IV, WAIS-5 or Stanford-Binet 5, given by a psychologist. Expect a report with a confidence interval of about five points either way, so roughly 133 to 143. That range stays above Mensa's line; our Mensa guide explains how to submit it. Our free IQ test is a practice screener, not a clinical score; here is why online results vary.
Other scores in this range
- IQ 136: about 1 in 122.
- IQ 139: one point from the fabled 140 line.
- IQ 140: Terman's selection line, about 1 in 261.
Frequently asked questions
Is 138 a high IQ?
Yes. A 138 is higher than about 99.4 percent of people and falls in the top classification band of both the Wechsler and Stanford-Binet scales. Only about 1 person in 177 scores this high or higher on a scale with standard deviation 15.
What percentile is an IQ of 138?
An IQ of 138 sits at about the 99.44th percentile. Roughly 0.565 percent of people score at or above it.
Would a 138 have qualified for Terman's gifted study?
Probably not. Terman's main selection line was an IQ of 140 on the Stanford-Binet of his era, which used ratio scores rather than today's deviation scale. The comparison is loose, because the tests and scoring methods differ.
Can I join Mensa with an IQ of 138?
Yes, if the score comes from an accepted test taken under supervision. Mensa admits people at or above the 98th percentile, which is 130 on Wechsler scales. American Mensa accepts about 200 standardized tests.
Does an IQ of 138 guarantee success?
No. It predicts better average outcomes in school and work, but the effects are moderate. Terman's high scorers did well as a group, yet none won a Nobel, while two boys who fell short of his cutoff later did.
Sources
- Genetic Studies of Genius — 1921 start, 1,528 children, college completion vs 8 percent
- Simonton, D. K. (2018). Your IQ Matters Less Than You Think. Nautilus — Average IQ 151; Shockley and Alvarez screened out; no Termite Nobel
- Warne, R. (2020). Terman's non-geniuses: Shockley and Alvarez — 140 selection line; why the misses were expected
- Deary, I. J., et al. (2007). Intelligence and educational achievement. Intelligence, 35(1) — 0.81 correlation, 70,000+ students
- Strenze, T. (2007). Intelligence and socioeconomic success. Intelligence, 35(5) — Meta-analytic correlations with education, occupation, income
- IQ classification — Wechsler bands
- Western New York Mensa: How to join — About 200 accepted tests
Editorial standard: IQ figures for historical figures are retrospective estimates and are labeled as such; scores for living people appear only when the person disclosed them or a reputable outlet reported them. Corrections: contact the editors.
