What Is a Good IQ Score? The Honest Answer, Band by Band

Good compared with whom

An IQ score only means something against a reference group. Modern tests set that group's middle at 100 and its standard deviation at 15, so every score converts directly into a percentile: the share of same-age peers who scored lower.

The arithmetic is unforgiving at the top. A 110 beats about 75 percent of people. A 115 beats about 84 percent. A 120 beats about 91 percent, a 125 about 95 percent and a 130 about 97.7 percent. Each step of five points above 115 cuts the remaining field roughly in half.

By that measure, any score above 110 is good in the plain sense of the word: higher than most. Our IQ percentile calculator converts any score, and what is the average IQ explains why 100 is the anchor.

Every IQ band, from Extremely Low to Very Superior

The labels come from the test publisher. The table uses the Wechsler adult scale (WAIS-IV) and the Wechsler children's scale (WISC-V), the two most widely used individual tests. Shares are theoretical proportions on a normal curve.

No single naming scheme governs every publisher, which is why a psychologist reports the test used and follows its terms. The full cross-test chart is in IQ scale and classification; the landmark scores are IQ 110, IQ 120 and IQ 130.

What an IQ score predicts

School is where IQ predicts best. Ian Deary and colleagues followed more than 70,000 English children in a study published in Intelligence in 2007. General intelligence measured at 11 correlated 0.81 at the latent level with national exam results at 16. Among children at the average level of ability, 58 percent earned five or more good GCSE grades; one standard deviation above average, the figure was 91 percent.

Life outcomes follow at lower strength. Tarmo Strenze's 2007 meta-analysis of longitudinal studies found corrected correlations of 0.56 between intelligence and educational attainment, 0.43 with occupational status and 0.20 with income. A 2011 meta-analysis led by Catherine Calvin, pooling 16 studies and more than 1.1 million people, found that each standard deviation of higher test scores in youth was associated with a 24 percent lower risk of death over the follow-up period.

These are population tendencies. They describe averages across thousands of people, not destinies for one.

What a score does not settle

The workplace evidence has been revised downward. For two decades, cognitive ability was described as the single best predictor of job performance. In 2022, Paul Sackett, Charlene Zhang, Christopher Berry and Filip Lievens re-examined the meta-analyses in the Journal of Applied Psychology and argued that standard statistical corrections had inflated the figures. Their estimate for cognitive ability was 0.31, behind structured interviews at 0.42, job knowledge tests at 0.40, empirically keyed biographical data at 0.38 and work samples at 0.33.

Strenze reached a similar verdict on success. Intelligence, he wrote, is "not an overwhelmingly better predictor than parental SES or grades." A correlation of 0.20 with income means IQ accounts for about 4 percent of the variation in earnings. Effort, health, opportunity, personality and luck share the rest.

A good score, in short, shifts the odds. It does not write the result. For the abilities tests leave out, see emotional intelligence versus IQ.

Good enough for what

Some thresholds are fixed. Mensa admits people who score in the top 2 percent on an approved, supervised test: 130 on a scale with a standard deviation of 15, or 148 on the Cattell scale, which uses 24. The route is in how to join Mensa.

Other thresholds are myths. No professional body defines a "genius IQ." The 140 line that circulates online appears in no current Wechsler classification table; the case is set out in what is a high IQ. Occupations, meanwhile, recruit across wide ranges, and the published averages by job show heavy overlap; IQ by profession reviews the data.

How to read your own number

Start with the test. A score from a psychologist-administered Wechsler or Stanford-Binet carries a reported confidence interval, usually 95 percent, of roughly five points either way. A 118 is really a range from the low 110s to the low 120s. Children who retake the very same test usually gain five or six points from familiarity alone, so a second result is not automatically a truer one.

Check the norms, too. Every reputable test compares you with people of your own age, and a score earned on an edition normed decades ago tends to run high, because raw performance rose for most of the last century. A 125 on a current edition and a 125 on a 1980s edition are not the same claim. Ask which test, which edition and which year before comparing numbers with anyone else.

Online scores are weaker evidence. Most use unpublished norms and self-selected test takers. Our free IQ test is a practice screener built to show where you likely fall, not a diagnosis; are online IQ tests accurate explains the gap. More guides sit in our IQ guides library.

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.