Reference
Average IQ by age
Every age group has an average IQ of 100, because scores are calculated within age bands. What changes with age is raw performance — and it changes in two opposite directions at once.
Why every age group averages 100
IQ is a deviation score calculated within an age band. A 62-year-old is compared against other 62-year-olds, a 19-year-old against other 19-year-olds, and in each case the middle of that group is defined as 100.
So a table of “average IQ by age” showing different values for different ages is either using raw scores, or using a single set of norms across all ages, or made up. On any properly normed test, the answer is 100 for everyone.
That design choice is deliberate. The score is meant to express your standing among your peers, which is the question that matters for education, employment and clinical assessment. It is not meant to express absolute cognitive capability, and it does not.
A 70-year-old scoring 115 and a 25-year-old scoring 115 are both at the 84th percentile of their own age group. Their raw performance on the tasks is not the same, and the score is not trying to say that it is.
What raw performance actually does
Underneath the age norms, two things happen in opposite directions, and the distinction between them is the central finding in the cognitive ageing literature.
Fluid abilities — reasoning with novel material, processing speed, working memory — peak somewhere in the twenties and decline gradually thereafter. The decline is measurable from the thirties and accelerates after about sixty.
Crystallised abilities — vocabulary, general knowledge, accumulated expertise — rise through adulthood and hold up or continue improving into the sixties and seventies, often declining only in the final years of life.
This is why an older adult can outperform a younger one comprehensively on a vocabulary subtest while scoring lower on a matrix reasoning subtest, and end up with a similar full-scale score once the age norms are applied.
| Age range | Fluid abilities | Crystallised abilities |
|---|---|---|
| 16–20 | Approaching peak | Rising quickly |
| 20–30 | Peak | Rising |
| 30–45 | Slow decline begins | Still rising |
| 45–60 | Steady decline | Plateau or slow rise |
| 60–75 | Faster decline | Held, slow decline late |
| 75+ | Marked decline | Decline, often accelerating |
Children and the stability problem
Childhood IQ scores are less stable than adult ones, and the earlier the test the less stable the result.
Scores before about age six correlate only weakly with adult outcomes, and clinical guidance is consistent that a single preschool score should not be treated as predictive. Stability improves through later childhood and is reasonably good from adolescence onwards.
This matters for gifted identification, where a decision is often made on a single test taken young. A score with a five-point confidence interval, taken at an age when scores are least stable, applied against a hard administrative threshold, is sorting partly on noise. More on this.
Why old age-comparison tables are misleading
Cross-sectional studies — comparing 70-year-olds with 30-year-olds today — overstate cognitive decline, for a reason that took the field decades to work out.
Today's 70-year-olds grew up with less schooling, different curricula and different exposure to abstract reasoning than today's 30-year-olds. The Flynn effect means each cohort scored higher than the one before at the same age. A cross-sectional comparison therefore confounds ageing with cohort.
Longitudinal studies, which follow the same people over time, consistently find smaller declines than cross-sectional ones. The truth is somewhere in between: decline is real, and it is less steep than the cross-sectional picture suggests.
The mental-age tables you may have seen
Occasionally a table circulates showing average IQ rising through childhood and falling through adulthood. That table is using the original ratio formula — mental age divided by chronological age — which was abandoned in the 1930s.
The ratio breaks for adults because mental age plateaus in adolescence while chronological age keeps going. Applied literally it would make every adult's IQ decline steadily from about twenty, which is an artefact of the arithmetic and not a finding about people.
Every modern test uses the deviation score instead. How the calculation actually works.
Questions
Questions people ask
What is the average IQ by age?
100 for every age group. IQ is calculated within age bands, so the middle of each age group is defined as 100 by construction.
At what age does IQ peak?
Raw performance on fluid reasoning tasks peaks in the twenties, while vocabulary and accumulated knowledge continue rising into the sixties. The age-normed score does not peak, because it is relative to your own age group.
Does IQ decline with age?
Raw fluid ability declines gradually from the thirties and faster after sixty; crystallised ability holds up much longer. The age-normed score stays centred on 100 for every age group.
Is a child's IQ score stable?
Not before about age six, where scores correlate only weakly with later outcomes. Stability improves through later childhood and is reasonably good from adolescence.
Why do cross-sectional studies show steeper decline?
Because they confound ageing with cohort effects. Older cohorts were educated differently and scored lower at every age, so comparing age groups today overstates the decline.
Find out where you land
Ten reasoning questions, an instant band and an explanation for every answer.