What does an IQ of 140 mean?

An IQ of 140 falls in the very superior band of the standard scale, at approximately the 99th percentile — as high as or higher than about 99% of the population. About 1 in 261 people score 140 or higher. Here is what that actually means, and just as importantly, what it does not.

The score in context

IQ scores sit on a scale constructed with a mean of 100 and a standard deviation of 15. A score of 140 therefore sits 2.67 standard deviations from the population average. About 68% of people score between 85 and 115, and 95% between 70 and 130 — the further from 100 a score sits, the faster the surrounding population thins out.

Scored close to 140 rather than exactly on it? Anything from 138 to 142 reads essentially identically: the gap is inside a single test's margin of error, so everything on this page — the classification, the percentile region, and the interpretation — applies across an IQ of 138, 139, 140, 141 or 142 alike.

What this band looks like in practice

Scores of 130 and above sit in the very superior band — two full standard deviations past the mean, reached by roughly two people in a hundred. This is the conventional threshold for high-IQ society membership and the level usually meant by the word "gifted". In a school year of two hundred pupils, perhaps four sit here.

At this level, tests themselves start to strain. Ceiling effects appear: a test built for the general population contains few items hard enough to distinguish 130 from 145, so scores at the top are measured with more error, not less. Extraordinary claims about very high scores — anything above 150 or so — should be read with the knowledge that almost no test taken at scale is calibrated out there, ours included.

The lived reality of this band is more mixed than the mythology suggests. The measurable advantages are real: rapid learning, ease with deep abstraction, and comfort in intellectually dense fields. But large studies following high scorers across decades find their life outcomes driven by the same unglamorous factors as everyone else's — conscientiousness, mental health, opportunity, and luck. A score here is a genuine and rare cognitive resource; what it becomes is a biography question, not a psychometric one.

Playing to the strengths of this band

At two standard deviations and beyond, the practical challenges invert. Capability is rarely the binding constraint; finding problems worthy of it is. People in this band report a consistent pattern — under-stimulation in general-pace environments, a tendency to intellectualise problems that need action, and difficulty calibrating communication for audiences who do not share their speed. The strategies that show up among those who convert this band into a satisfying life are concrete: seek out the few environments genuinely paced for it (research, deep technical work, competitive fields), build the translation skill of making complex things clear to any audience, and treat persistence as a trained discipline rather than assuming ability substitutes for it — because across every longitudinal study of high scorers, it never has.

How 140 compares with the anchor scores

The standard deviation anchors — 70, 85, 100, 115, 130, 145 — are the reference points every chart is built on. Here is 140 placed among them.

How much to trust a score of 140

Every IQ measurement carries error. Professionally administered tests report a confidence band of several points around any score, and single online sittings are noisier still: fatigue, distraction, rushing, or unfamiliarity with matrix puzzles can move a result meaningfully in either direction. The sensible reading of a 140 is "somewhere in the very superior region, give or take" — a good estimate, not a serial number.

The percentile is the part worth remembering. Scores can be rescaled; the fact of performing at roughly the 99th percentile of the comparison population is the stable, plain-English content of the result. When comparing results across different tests, always compare percentiles, never raw numbers, because different tests norm against different populations.

140 on other scales

Different tests have historically used different standard deviations, so the same underlying performance carries different numbers on different scales. A modern SD 15 score of 140 corresponds to approximately 143 on the older SD 16 Stanford-Binet convention and 164 on the SD 24 Cattell scale — identical percentile, different arithmetic. This is why comparing raw numbers across tests misleads, and why every comparison should run through percentiles.

Getting the most from a matrix reasoning test

If you plan to test or retest, a few things measurably help performance without gaming anything: take it rested and undistracted, since working memory is the first casualty of fatigue; read each grid row by row and column by column, naming the rule to yourself before looking at the options; and manage the clock — early questions should take seconds, banking time for the multi-rule items at the end where the discrimination between scores actually happens. Guessing carries no penalty on this test, so never leave a question unanswered as time runs out. And space any retest by weeks rather than hours: immediate retakes inflate scores through memory of specific items, which tells you nothing new.

What a single number leaves out

One score compresses many abilities. Verbal reasoning, spatial visualisation, working memory, and processing speed correlate, but imperfectly — two people at 140 can have quite different cognitive profiles, and most people have a spread of 10 to 20 points between their strongest and weakest domains. A matrix reasoning score specifically targets the fluid, pattern-based core of ability that correlates most strongly with the general factor, which is why single-format tests of it are meaningful — but no single number is a full portrait of a mind, at 140 or anywhere else.

Frequently asked questions