What Does an IQ of 170 Mean? How Rare It Is and What Tests Can Show

An IQ of 170 is 4.67 standard deviations above the average of 100 on the standard Wechsler scale, a score that about 1 person in 653,000 would reach, yet the main adult tests stop at 160 and cannot print it. The 170 that appears in books, quizzes and headlines is usually a child’s ratio IQ from an old test, a score on a wide-spread scale such as Cattell’s, an unsupervised online result or an estimate. This guide covers what 170 means on each scale, how it compares with the one-in-a-million line of the Mega Society, what Lewis Terman found among the children who scored this high in the 1920s, how accurate any single 170 can be, and what to do when someone quotes you one.

• Rarity: 1 in 653,000 on a standard deviation (SD) 15 scale, roughly 12,600 people among 8.2 billion if the bell curve held that far out, which nobody has been able to check.
• Other scales: the same printed 170 is 1 in 165,000 on SD 16 and only 1 in 565 on Cattell’s SD 24.
• Ceiling: the WAIS-IV and Stanford-Binet 5 stop near 160. Pearson’s WISC-V extended norms reach 210, for children only.
• Evidence: 77 of the 1,070 children in Terman’s study who sat the Stanford-Binet scored 170 or more, on a 1916 ratio scale that is not a modern deviation IQ.

How rare is an IQ of 170?

On a normal curve with a mean of 100 and a standard deviation of 15, a score of 170 sits at z = +4.67. About 0.00015% of people would reach it, which is 1 in 653,000. Applied to a world population of about 8.2 billion that is roughly 12,600 people, spread across every age and country. The table shows how steeply the numbers fall from 150 to 175.

Two cautions go with those numbers. The bell curve is a model: a standardisation sample of 3,000 people is expected to hold about 0.005 of a person at 170 or above, so the top rows of every norm table are extrapolated rather than observed (see how rare is a high IQ score?). And almost none of the people the curve implies have ever sat a test that could measure them. To place any score on the curve, use the IQ percentile calculator; the IQ bell curve explains why anything past about 145 is extrapolation. At 160 the figure is 1 in 31,600, covered in what an IQ of 160 means, and at 180 it is 1 in 20.7 million, covered in what an IQ of 180 means.

What is an IQ of 170 on other IQ scales?

A bare “170” hides which scale was used. The Wechsler tests and the current Stanford-Binet use a standard deviation of 15, the older Stanford-Binet Form L-M used 16, and Cattell’s scales use 24. Each point above 100 is worth fewer standard deviations on a wider scale, so the same number is far less rare.

So a Cattell 170 is roughly a 144 on the Wechsler scale, and an L-M 170 is roughly a 166. That is one reason numbers such as 170 turn up in older and high-range literature: they were quoted on wide scales or as children’s ratio scores. Standard deviation 15 versus 16 covers the mechanics, and the IQ score converter moves a score between scales.

Is an IQ of 170 the Mega Society line?

Close, if you take that society’s line as a percentile. High-IQ societies set their bars as rarities, not as point scores, and the Mega Society’s is one person in a million. Converting each line to the SD 15 and SD 16 scales shows where 170 falls on the ladder. The Mega figures are our arithmetic; the society itself publishes a percentile, and Beyond Mensa explains why the numbers get shakier the further out the line is drawn.

A 170 on SD 15 is 1 in 653,000, slightly less rare than the one-in-a-million line, and a 160 is the Prometheus line of one in 30,000. Two things limit the comparison. The tests these societies accept were not normed on samples large enough to place anyone that far out, and the Mega Test was taken unsupervised and untimed, which is why scores after 1994 stopped being accepted.

Can any test report an IQ of 170?

For adults, no. The WAIS-IV and the Stanford-Binet Fifth Edition cap their Full Scale IQ near 160, so a report that prints 160 means “at or above the top of what this test can measure”, and a report that prints 170 from either should be questioned. The IQ ceiling effect and what is the highest IQ score a test can report? explain the two separate limits, the rarity of the score and the end of the test’s item list.

For children there is one genuine route. Pearson’s WISC-V Extended Norms (Technical Report 6, 2019) raise the composite maximum to 210 for children who reach the ceiling on several subtests, so a child can legitimately be reported at 170. The extension belongs to that one children’s test. An online quiz can print any number it likes but cannot certify one this rare, and a result above about 145 from that kind of test says more about the scoring formula than about the test-taker. The same logic is why an IQ of 180 appears mostly as a child’s ratio score.

What did Terman find above 170?

The best evidence on children this high is a study that began a century ago. In 1921-22 Lewis Terman searched the public schools of Los Angeles, San Francisco, Oakland, Berkeley and Alameda, about 168,000 pupils, and followed 1,528 children who scored in the top 1% on IQ tests (see our data study of the Termites). Of the 1,070 tested on the individual Stanford-Binet, the average was 151.0, the range ran from 135 to 200, and 77 scored 170 or more. In the main group of 643 children, Terman’s corrected scores put 43 at 170 or above: 22 boys and 21 girls, and the three highest scores were girls’.

The pattern is the point. A normal curve with SD 15 predicts about 0.26 pupils at 170 or above among 168,000, and SD 16 about 1.0; the search found 43, and not every pupil was tested, so the true count was higher. Two things explain most of the gap. The 1916 scale gave ratio IQs, which stretch in the tails (a child with a mental age of 8.5 at the age of 5 scores 170 whatever the curve says), and the sample was a hand-picked, teacher-nominated group, not a random draw. So 77 is not a count of people who would score 170 on today’s tests; see deviation IQ versus ratio IQ.

What happened next is less dramatic than the numbers suggest. When the Termites were tested again as adults in 1950-52 on a hard verbal test, the Concept Mastery Test, the 48 who had scored 170 or more as children averaged 155.8, against 146.2 for the 160 to 169 group, 136.5 for 150 to 159 and 131.8 for 140 to 149. The correlation between childhood score and adult result was only .29, and the spread inside each group was wide (a standard deviation of 23.1 in the top group). Their own children averaged 132.7, well above 100 but far below the parents’ 151, which is regression to the mean in action (see regression to the mean). No Termite is known to have won a Nobel Prize, while William Shockley, a future Nobel physicist, was tested as a boy, scored 129 and 125, and missed the cut-off.

How accurate would a score of 170 be?

Every score carries measurement error, and an extreme score is more likely than a typical one to have been helped by chance. Kelley’s formula corrects for it: the expected true score is 100 plus the test’s reliability times the distance of the observed score from 100. The table applies it to an observed 170.

The person who scores 170 is, on average, a little less exceptional than the score says, and a retest on a different instrument usually lands lower. The table is also optimistic, because error grows toward the tails and no test has enough hard items at this level. Outdated norms add to it: scores inflate by roughly three points a decade as norms age (why IQ norms expire), so a 170 on a 20-year-old norm sample may be nearer 164 on a current one. And practice matters: practice effects raise scores on repeat sittings. The margin-of-error explainer shows the same arithmetic nearer the middle of the scale. Read a 170 as “far into the top fraction of a percent”, not as a point.

Is an IQ of 170 genius level?

Genius is not a psychometric category. Test manuals label scores of 130 and above “very superior” or “extremely high” and stop there, and Terman’s 1916 scheme labelled 140 and above “near genius or genius”. Work that changes a field depends on creativity, opportunity and years of effort, which is why a score alone predicts it poorly; the long-run data on mathematically precocious youth is summarised in does IQ above 120 stop mattering? and in what is a genius IQ score. For the score most often attached to the word, see is 130 IQ genius?.

Who has an IQ of 170?

No verified list exists. Most figures attached to famous people are retrospective estimates from biographies, not test scores, as where celebrity IQ numbers come from explains, and the few children measured this high were measured on ratio scales. Our celebrity IQ profiles label each number by how it was produced, and the highest IQ ever recorded shows how headline figures such as 228 arose. Treat any list headed “people with an IQ of 170” as a list of claims.

If you have been told you scored 170

• Find the scale. Ask which test it was and whether its spread was 15, 16 or 24. The same number means very different things.
• Ask whether it is a ratio score. A child’s 170 from an old Stanford-Binet is mental age over age, not a rank among same-age peers.
• Check who gave it. An individually administered test from a psychologist carries weight; an unsupervised web quiz does not.
• Look for the ceiling. If the test stops at 160, a 170 cannot come from it.
• Quote a percentile or a range. “Far into the top fraction of a percent” is honest; “exactly 170” is not. At the other end of the curve the same caution applies: an IQ of 65 is measured with far more precision than a 170.

Quick answers

• Is an IQ of 170 good? It is far above any practical threshold: 4.67 standard deviations above average on SD 15.
• What percentile is an IQ of 170? About the 99.9998th on SD 15.
• How many people have an IQ of 170? Roughly 12,600 of 8.2 billion if the bell curve held, which nobody has been able to verify.
• Is 170 higher than 160? Yes, but 160 is the printed top of the main adult tests, so a 170 needs a different instrument and a stronger caveat.
• Is an IQ of 170 possible? On paper, yes. As a child’s ratio IQ, yes: Terman found 77. As a certified adult score on a current test, no.
• Can an online test give a 170? It can print one. It cannot certify one.

The bottom line

An IQ of 170 means 4.67 standard deviations above average on the Wechsler scale: about 1 person in 653,000, beyond the ceiling of the main adult tests and just under the one-in-a-million line the Mega Society draws. Most 170s in circulation are children’s ratio IQs, wide-scale scores or estimates, and Terman’s 77 children show how a 1916 scale stretched in the tail. The number that travels between tests is the percentile, so ask for the scale, the test and the date before the figure is taken at face value. To see where your own score sits, start with the IQ Metrics IQ test and the IQ tests hub.

Sources

• Terman, L.M. et al. (1925): Genetic Studies of Genius, Volume I: Mental and Physical Traits of a Thousand Gifted Children, Stanford University Press (Table 5 for the corrected score bands; the search of about 168,000 pupils), and Terman, L.M. and Oden, M.H. (1959): Volume V: The Gifted Group at Mid-Life (the 77 children at 170 or more; the Concept Mastery Test results). Both as tabulated in our Terman study page.
• Pearson (2019): WISC-V Technical Report 6, Extended Norms.
• Terman, L.M. (1916): The Measurement of Intelligence, Houghton Mifflin, the classification table that set the near-genius line at 140.
• Rarity, percentile, society-line and Kelley figures are computed from the normal distribution. The scale facts, ceilings and society thresholds follow our own guides to an IQ of 180, the ceiling effect and high-IQ societies.