The IQ scale is a standardized scoring system with an average of 100 and a standard deviation of 15, dividing scores into classifications from Extremely Low (below 70) to Very Superior (130 and above). It is not a ruler that measures a fixed quantity of intelligence; it is a ranking system that tells you where a person's test performance sits relative to everyone else's. This guide walks through the full scale, the bell curve behind it, how the scale came to be, and what different tests actually measure.
The full IQ scale and classification labels
Most modern tests use classification bands modeled on the Wechsler scales, the most widely administered IQ tests in the world:
| IQ range | Classification | Share of population |
|---|---|---|
| 130 and above | Very Superior | ~2.1% |
| 120–129 | Superior | ~6.4% |
| 110–119 | High Average | ~15.7% |
| 90–109 | Average | ~50% |
| 80–89 | Low Average | ~15.7% |
| 70–79 | Borderline | ~6.4% |
| Below 70 | Extremely Low | ~2.1% |
The labels are conventions, not diagnoses — a score in the Borderline band does not by itself diagnose anything, and a Very Superior score does not certify genius. Newer editions of the Wechsler tests have softened some labels ("Extremely High" instead of "Very Superior"), but the score bands remain the same. For a version of this table with percentiles attached, see our IQ score chart, and for what counts as a strong score in practice, see what is a good IQ.
The bell curve: why the scale looks the way it does
IQ scores follow a normal distribution — the familiar bell curve. Scores cluster densely around the mean of 100 and thin out symmetrically toward both tails. The standard deviation of 15 is the curve's unit of distance, and it produces three numbers worth memorizing:
- About 68% of people score within one standard deviation of the mean — between 85 and 115.
- About 95% score within two standard deviations — between 70 and 130.
- About 99.7% score within three — between 55 and 145.
This is why extreme scores are so rare: only about 1 person in 50 scores 130 or above, and only about 1 in 1,000 reaches 145. It is also why every IQ score maps cleanly onto a percentile — a rank against the population — which you can look up in our IQ percentile table.
A short history of the IQ scale
Binet: the first practical test
The scale begins in Paris in 1905, when Alfred Binet and Théodore Simon built a test to identify schoolchildren who needed extra help. Their innovation was the concept of mental age: a child who solved problems typical of an average 10-year-old had a mental age of 10, whatever their birth certificate said.
Stanford-Binet and the ratio IQ
Stanford psychologist Lewis Terman adapted the test for the United States in 1916 as the Stanford-Binet. It used German psychologist William Stern's formula for an "intelligence quotient": mental age divided by chronological age, multiplied by 100. A 10-year-old performing like a typical 12-year-old scored 120. The term "IQ" comes from this literal quotient.
Wechsler and the deviation IQ
The ratio formula breaks down for adults — mental age plateaus, so dividing by chronological age becomes meaningless. David Wechsler's solution, introduced with his 1939 test and universal today, was the deviation IQ: compare each person to others of the same age and express the result as a position on a normal curve with mean 100 and SD 15. Modern "IQ" is therefore not a quotient at all — it is a standardized rank. This age-norming is also why average IQ stays 100 at every age even as raw abilities change across life.
One practical footnote: because tests are periodically re-normed and raw performance rose through the 20th century (the Flynn effect), today's 100 reflects a higher raw performance than the 100 of 1950 — see our guide to the average IQ for details.
What different tests measure
"IQ" is a family of tests, not a single instrument, and they do not all probe the same abilities:
- Wechsler scales (WAIS for adults, WISC for children) combine four index areas — verbal comprehension, perceptual reasoning, working memory and processing speed — into a full-scale score.
- Stanford-Binet (5th edition) measures five factors, including fluid reasoning, knowledge, quantitative reasoning, visual-spatial processing and working memory.
- Raven's Progressive Matrices is entirely nonverbal — abstract pattern completion — making it popular for testing across languages and cultures.
- Cattell Culture Fair similarly minimizes language, but reports scores on an SD-24 scale, so its numbers run higher for the same percentile.
That last point matters when comparing scores: a 148 on the Cattell scale is the same rank as a 130 on a Wechsler test. Percentile, not the raw number, is the common currency.
Despite their different formats, well-constructed tests correlate strongly with one another, which is the empirical basis for the idea of a general ability factor ("g") running through them. But the correlation is not perfect — a person can plausibly land ten points apart on two different instruments taken in the same month. That is normal, expected behavior of the scale, not evidence that one test "lied." It also means the scale has a practical ceiling: above roughly 145–150, too few people exist in any norming sample for standard tests to rank scores reliably, which is why extraordinary claims about record scores rest on shaky ground — a story we unpack in the highest IQ ever recorded.
What each band means in plain English
Labels are shorthand; here is the practical reading of each band, keeping in mind that the borders are soft and individuals vary widely within every range:
- Very Superior (130+): about 1 person in 50. The usual threshold for gifted programs and high-IQ societies; abstract reasoning tasks that most people find hard tend to feel routine.
- Superior (120–129): roughly the 91st–98th percentile, a range common among people in cognitively demanding professions.
- High Average (110–119): comfortably above the mean — about one in six people. Typically associated with smooth progress through higher education.
- Average (90–109): the broad middle of the curve, holding half the population. Most jobs, degrees and life outcomes are well within reach across this band.
- Low Average (80–89): below the midpoint but within the normal range; about one in six people score here.
- Borderline (70–79): a range where standardized academic work is often harder; clinicians look at adaptive functioning, never the score alone.
- Extremely Low (below 70): about 2% of scorers. A score here is one criterion considered in assessments of intellectual disability, always alongside real-world functioning.
Reading your own score on the scale
Three cautions keep the scale in perspective. First, every score carries measurement error — a typical confidence interval is about ±5 points, so treat classification borders as fuzzy, not sharp. Second, online tests provide estimates for self-knowledge and practice; they are not clinical instruments, and formal decisions (educational placement, disability assessment) require a professionally administered test. Third, the composite number summarizes several distinct abilities — two people at 105 can have very different cognitive profiles. The scale tells you where you rank; a good assessment also tells you how you think.