What Does an IQ of 70 Mean?

An IQ of 70 is at the 2.3rd percentile: about 2.3% of people of the same age score at this level or lower, and 97.7% score higher. On the WAIS-IV it is classed as Borderline. It is an estimate with a margin of error of several points, and it is never, on its own, a diagnosis.

IQ 70 percentile and how it is calculated

IQ tests are scaled so that scores follow an approximately normal distribution with a mean of 100 and a standard deviation (SD) of 15. A score of 70 is 2 standard deviations below the mean (z = -2.000). Plugging that into the normal cumulative distribution function gives a percentile of 2.3. Put simply: in a representative group of 1,000 people of your age, about 23 would score 70 or lower.

Some reports use a slightly different convention — for example the percentage of people scoring strictly below 70, or a value read from the test’s own norm tables rather than the ideal curve — so you may see a percentile a fraction of a point different. The full list for every score from 55 to 160 is in our IQ percentile chart.

How 70 compares with nearby scores

Is an IQ of 70 low?

It is a low score — about 2% of people score this low or lower — and on its own it is not enough to diagnose anything. See also what counts as a good IQ score.

How rare is a 70 IQ?

On the normal model, about 1 in 44 people scores 70 or lower — about 23 per thousand people. In a school of 1,200 pupils that would be around 27 students; in a city of one million, roughly 22,800 people. A score in this range is therefore uncommon but far from exceptional.

70 IQ classification

Different tests use different descriptive words for the same number. For a full-scale score of 70:

These labels describe where a score falls relative to peers; they are not diagnoses. Newer reports, such as the WAIS-5 sample report, use plainer wording (“Average”, “Above average”, “Very high”). See all bands in IQ score ranges and classification.

What an IQ of 70 means in practice

A score of 70 sits exactly two standard deviations below the mean, which is why it became the traditional reference point for intellectual disability. Today the cut-off is not a hard line. The DSM-5 (and DSM-5-TR) defines intellectual disability by three things together: deficits in intellectual functioning (typically a score around 65–75 once measurement error is allowed for), deficits in adaptive functioning (everyday conceptual, social and practical skills), and onset during the developmental period. A person who scores 70 but manages work, money, relationships and self-care independently would not meet the definition.

The same caution appears in law. In Hall v. Florida (2014) the US Supreme Court rejected a rigid “IQ 70 or below” rule, holding that courts must take the test’s standard error of measurement into account. In other words, a measured 70 and a measured 72 are not meaningfully different numbers.

If a child or adult has scored around 70, the useful next step is a full assessment of adaptive behaviour (for example with the Vineland or ABAS scales) alongside the IQ test. That combination opens access to school support plans, disability services and vocational programmes that are designed for exactly this profile.

Education, work and an IQ of 70: what research says

Large meta-analyses agree that IQ predicts educational and occupational outcomes on average, with plenty of individual exceptions. Strenze (2007), pooling longitudinal studies, found corrected correlations of about .56 between intelligence and education, .43 with occupational level and .20 with income. For job performance, Schmidt and Hunter (1998) estimated a validity of about .51 for general mental ability; a 2022 re-analysis by Sackett and colleagues revised it to about .31, still useful but leaving most differences in performance to other factors such as experience, conscientiousness and job knowledge.

How does 70 look against real occupations? The table uses Wolfram’s (2023) estimates from the UK Understanding Society survey and assumes a typical spread of about 13 points inside each occupation (the study found only about a quarter of IQ variance lies between occupations). It shows how many people in each group would be expected to score below 70 — a reminder of how much occupations overlap.

Read more in average IQ by occupation. These are group estimates from a short test battery, not entry requirements.

Measurement error: how precise is a score of 70?

A full-length clinical test such as the WAIS-IV or WAIS-5 has very high reliability (internal consistency around .97–.98 for the full-scale IQ), which still leaves a standard error of measurement (SEM) of about 2–3 points; Pearson’s WAIS-5 sample report lists an FSIQ SEM of 3.00. A 95% confidence interval is roughly ±1.96 SEM, so a measured 70 is best read as about 64 to 76. Shorter tests have larger errors (SEM 4–5 or more), widening the band to roughly 61–79.

Extreme scores also tend to move toward the mean on re-testing (regression to the mean). With a reliability of .96, the best estimate of the “true” score behind an observed 70 is about 71.2, which is why many reports centre the confidence interval slightly above the obtained score.

What an IQ of 70 does not tell you

• It does not measure creativity, practical wisdom, emotional skills, motivation or character.
• It is a sample of performance on one day; sleep, stress, language and familiarity with tests all affect it.
• Index scores (verbal, spatial, reasoning, memory, speed) can differ widely behind the same full-scale number.
• It does not fix the future: correlations of .2 to .6 leave room for very different outcomes among people with the same score.
• It cannot be compared across very different tests or eras without adjusting for norms.

The Flynn effect and old norms

The Flynn effect is the long-run rise in raw test performance across generations — roughly 3 IQ points per decade in 20th-century data (Trahan et al., 2014, estimated about 2.9). Its practical consequence is that a test scored against old norms overstates IQ. If your test was normed 10 years before you took it, a 70 could be inflated by around 3 points compared with current norms.

Online IQ tests and a score of 70

An online score of 70 is not the same as a clinical 70. Online tests usually lack a proper norm sample, can be retaken until a result looks good, and vary wildly in difficulty. Some research-grade public tools exist (for example the ICAR item bank; Condon & Revelle, 2014), but even they are designed for group research, not individual diagnosis. More in our guide to online IQ test accuracy.

How to get a proper IQ assessment

A valid IQ score comes from an individually administered, standardized test (for example the WAIS-5/WAIS-IV for adults, WISC-V for children, or Stanford-Binet 5) given by a licensed or registered psychologist. Expect 60–120 minutes of testing, a written report with index scores and confidence intervals, and a fee that varies by country and provider. Schools, university disability services, and clinical or neuropsychology services can refer you. Ask which test and which norms were used, and for the confidence interval, not just the single number.