What is an average IQ?

Last reviewed 6 September 2026

Why the average is exactly 100

Because somebody decided it should be. That is not a flippant answer - it is the whole explanation.

When a test publisher builds an IQ test, they administer it to a large standardisation sample selected to match a target population on age, sex, education, region and other characteristics. Then they take the raw score distribution from that sample and transform it so that the mean becomes 100 and the standard deviation becomes 15.¹

So "the average IQ is 100" is not a finding about humanity. It is a description of the scale. If everyone in the world woke up tomorrow able to solve twice as many matrix problems, the average IQ would still be 100 the moment the test was renormed. What would change is the raw performance behind the number.

What "average" covers

"Average" in everyday speech means one point. In test interpretation it means a band, and a wide one. The conventional average range is 90 to 109, which contains roughly half of all people. Widen it to 85-115 - one standard deviation either side - and you have captured about 68%.

Which means: most people are average, and the differences within that band are small enough that measurement error swamps them on any short test. Someone reported at 104 and someone reported at 97 have not been meaningfully distinguished.

Renorming and the Flynn effect

Here is where the "average is 100 by definition" point stops being a technicality.

Across the twentieth century, raw performance on IQ tests rose steadily. James Flynn documented the pattern across fourteen nations in 1987, and it now carries his name.² The largest meta-analysis to date pooled 271 independent samples covering close to four million people in 31 countries between 1909 and 2013. It put the gain at roughly 0.28 IQ points per year for full-scale scores - and notably larger for fluid reasoning (about 0.41 points a year) than for crystallised knowledge (about 0.21).³

Compounded over decades that is enormous. Someone scoring at the average on a 1950 norm set would score well below average against a 2010 norm set, without having got any worse at anything. Publishers respond by renorming every decade or two, which pulls the mean back to 100.

Two practical consequences:

• Old scores inflate. A score from a test normed thirty years ago is generally too high by modern standards. This matters in any setting where a threshold has legal or educational consequences.
• The gains are not universal or permanent. Analysis of Norwegian conscript data found the rise reversing in cohorts born after about 1975, and the pattern within families indicated environmental rather than genetic causes.⁴

Does average IQ change with age?

On a properly normed test, no - and that is deliberate. Scores are age-normed, meaning you are compared with people of your own age. A 70-year-old scoring 100 performed averagely for a 70-year-old.

Underneath that normalisation, raw performance does change with age, and the two broad abilities move differently. Fluid reasoning - solving novel problems - peaks in the twenties and declines gradually thereafter. Crystallised knowledge - vocabulary, accumulated information - holds steady or keeps rising well into later life.⁵ Age-norming hides this divergence by design, which is usually what you want and occasionally not.

One consequence for this site specifically: the test here is not age-normed, because doing so would require a standardisation sample it does not have. It leans on fluid reasoning items, so older test takers are likely to be scored slightly harshly relative to a properly age-normed instrument. That is stated in the limitations.

Average IQ by country: handle with care

Tables ranking countries by average IQ circulate widely. They deserve much more scepticism than they usually get, for reasons that are methodological before they are anything else.

• The underlying samples are often tiny and unrepresentative. Many entries in the best-known datasets rest on convenience samples - a few hundred schoolchildren in one city - extrapolated to an entire national population. Some entries are estimated from neighbouring countries rather than measured at all.⁶
• Tests are rarely normed for the population taking them. Administering an instrument standardised on one population to a very different one, in a second language, violates the assumptions that make the score interpretable.⁷
• The Flynn effect makes cross-era comparison invalid. Data gathered decades apart cannot be placed on one scale without correction, and often is not corrected.
• Schooling, nutrition, health and test familiarity all move scores. These vary enormously between countries and are precisely the environmental factors shown to drive the Flynn effect.⁴

The honest summary: differences in measured average performance between populations are real in the data, and the datasets themselves are too weak, and too confounded with schooling and measurement artefacts, to support the interpretations usually built on them.

Common questions

References

1. Wechsler, D. (2008). Wechsler Adult Intelligence Scale - Fourth Edition: Technical and Interpretive Manual. Pearson.
2. Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101(2), 171-191.
3. Pietschnig, J., & Voracek, M. (2015). One century of global IQ gains: A formal meta-analysis of the Flynn effect (1909-2013). Perspectives on Psychological Science, 10(3), 282-306. doi:10.1177/1745691615577701
4. Bratsberg, B., & Rogeberg, O. (2018). Flynn effect and its reversal are both environmentally caused. PNAS, 115(26), 6674-6678. doi:10.1073/pnas.1718793115
5. Salthouse, T. A. (2009). When does age-related cognitive decline begin? Neurobiology of Aging, 30(4), 507-514.
6. Wicherts, J. M., Borsboom, D., & Dolan, C. V. (2010). Why national IQs do not support evolutionary theories of intelligence. Personality and Individual Differences, 48(2), 91-96. doi:10.1016/j.paid.2009.05.028
7. American Educational Research Association, American Psychological Association, & National Council on Measurement in Education (2014). Standards for Educational and Psychological Testing. AERA.

Related reading