People search this for one of three reasons: they want to know where they’d stand, they’re worried their thinking has slipped with age, or they just took a free online test and want to know whether the number it gave them means anything.
All three deserve a better answer than the usual “it’s 100, here’s a bell curve.” The short version really is 100 — but the useful part is why, and what that does and doesn’t let you conclude.
The short answer
The average IQ for an adult is 100. With the standard deviation of 15 that virtually every modern test uses:
- About two-thirds of adults score between 85 and 115.
- About 95% score between 70 and 130.
- Roughly 1 in 6 adults scores above 115; roughly 1 in 6 scores below 85.
- Roughly 1 in 50 scores above 130 — the usual Mensa-eligibility line.
So “average” isn’t a narrow point; it’s a broad middle. An adult at 108 is, statistically, an ordinary adult. An adult at 88 is also closer to the middle of the distribution than to its edge.
Why 100 is a definition, not a finding
This is the part most articles skate past. The number 100 was not measured. It was assigned.
When a test publisher standardizes an IQ scale, they administer the test to a large, demographically representative sample, then take whatever raw-score distribution comes out and stretch and shift it so the mean lands on 100 and one standard deviation equals 15 points. Your IQ is your rank within that reference population, expressed on that fixed scale. The “average” is 100 the same way the freezing point of water is 0°C: it’s where someone decided to put the zero.
An IQ score isn’t a measurement of a substance — it’s a percentile rank dressed up in a familiar number. “The average adult has an IQ of 100” is true the way “the median income is the median income” is true. It’s the spread around 100, and what sits behind a given score, that carries the information.
Two consequences fall straight out of this and explain the rest of the article:
- Because it’s pinned to a reference population, “the average adult” depends on which population’s norms you’re standing on — and those have drifted over time (the section below).
- Because the score is also age-normed within adulthood, the average is 100 at 25 and at 65 — but the cognitive mix producing that 100 isn’t the same at both ages (the section after that).
The adult score bands
Here’s what the standard scale translates to in percentiles — the share of adults at or below that level — and the labels you’ll see on a report or in popular writing.
| IQ score | Percentile (among adults) | Typical label |
|---|---|---|
| 145+ | ~99.9th | Highly gifted |
| 130–144 | ~98th–99.9th | Gifted / very superior (Mensa range) |
| 120–129 | ~91st–97th | Superior |
| 110–119 | ~75th–90th | High average |
| 90–109 | ~25th–73rd | Average |
| 80–89 | ~9th–23rd | Low average |
| 70–79 | ~2nd–8th | Borderline |
| Below 70 | Below ~2nd | Extremely low |
And the standard caveat that applies to any real score: a well-built test still carries a margin of error of roughly ±5 points at 95% confidence. A reported 117 honestly means “about 112–122.” Read the band, not the point. For the longer treatment of what counts as “good” at each level, see what is a good IQ score.
The average adult of when?
If “100” is wherever the reference population’s middle was put, then a fair next question is: which reference population? And the answer matters more than you’d think, because raw cognitive test performance has been climbing for most of the last century — the Flynn effect.
Across many countries and many test types, raw scores rose by roughly 3 IQ-equivalent points per decade through the 20th century (Flynn, 1987; Pietschnig & Voracek, 2015). That’s why tests get re-standardized every decade or so — if they didn’t, the average would visibly creep above 100 and the scale would lose its meaning.
“The average adult has an IQ of 100” is only true relative to that test’s norming year. An adult scoring exactly average on a 1948 test would, by the raw performance behind that score, sit well below 100 against today’s norms — on the order of 15–20 points lower over that span. So the average adult’s score has been 100 the whole time, but the average adult’s raw performance on the same items has moved a lot. Comparing an IQ measured in 1960 with one measured today, point for point, is comparing two different yardsticks. (The gains have slowed or stalled in several wealthy countries since the 1990s — the so-called negative or “reverse” Flynn effect — which is itself an active research question.)
None of this makes IQ useless — within a properly normed test, the rank ordering it produces is one of the most stable and predictive measurements in psychology. It just means “the average adult IQ is 100” is a statement about a scale’s construction in a given era, not a fixed fact about human brains.
A 25-year-old’s 100 vs. a 55-year-old’s 100
IQ tests are age-normed across adulthood too, so the average is 100 at every adult age. But the cognitive ingredients that add up to that 100 shift, and this is where a lot of needless worry comes from.
Psychologists since Cattell and Horn have split adult intelligence into two strands:
- Fluid intelligence — reasoning with novel problems, pattern recognition, working memory, processing speed. Peaks in the late teens to mid-20s, then declines slowly — roughly 1–2 IQ-equivalent points per decade in raw terms, faster after 60 (Salthouse, 2009).
- Crystallized intelligence — vocabulary, general knowledge, verbal reasoning, learned expertise. Keeps rising into the 50s–70s, then plateaus.
So a 25-year-old’s average score leans on fast, novel reasoning; a 55-year-old’s average score leans on accumulated knowledge and well-rehearsed reasoning. Put both on the same composite IQ test and they both land near 100 — the curves are pulling in opposite directions and the composite splits the difference.
Noticing that you’re a beat slower on timed puzzles than you were at 22 is not evidence your IQ has dropped. Processing speed is the most age-sensitive piece of the battery, and it’s also the one most visible to you in daily life — while the vocabulary and judgment you’ve gained in the same years are quietly working in the other direction and are far less perceptible from the inside. A single composite number won’t show you either trend; a domain-by-domain breakdown will. If real, persistent cognitive change is the worry rather than the ordinary tempo of middle age, that’s a conversation for a physician, not a quiz.
For the full lifespan picture — childhood through late life — see average IQ by age.
Why online tests hand adults inflated numbers
If a free online test just told you your IQ is 124, here’s the uncomfortable part: that number is probably too high, and the reasons are structural, not personal.
- They’re rarely standardized on a representative sample. A proper IQ test is normed on a stratified national sample. Most online tests are normed — if at all — on whoever happened to take them, which is a self-selected, internet-using, curious-about-IQ crowd that skews above the population mean. Compare yourself to that group and “average” performance gets labeled “above average.”
- Flattery is good for business. A test that tells most takers they’re 90 gets shared less, and converts worse, than one that tells most takers they’re 125. There’s a quiet commercial gravity toward generous scoring, especially on tests that then try to sell you a certificate.
- Practice and format effects. Many online tests recycle a small pool of Raven’s-style matrix items that circulate widely; if you’ve seen the format before — and most people have — you’ll do better than a true first-exposure baseline.
A reasonable rule of thumb: an online IQ test is a decent way to get a rough, relative read on your reasoning across different domains, and a fun way to see your own profile of strengths. It is not a way to get a defensible number. If you need one of those — for a Mensa application, a clinical question, an accommodation — the route is an individually administered test like the WAIS-IV with a licensed psychologist. We build the most honest version of an online test we can, and that’s still what we’d tell you.
“Average adult IQ” is not “average national IQ”
One last source of confusion. Searching “average IQ” quickly surfaces tables ranking countries by “national IQ” — Country A is 105, Country B is 84, and so on. Those numbers are not the same kind of thing as “the average adult IQ is 100,” and treating them as comparable is a mistake.
“The average adult IQ is 100” is a near-tautology about how a single well-normed test is scaled within its reference population. The country-ranking figures come from a separate body of work (most associated with Richard Lynn) that has been heavily and repeatedly criticised by other researchers for relying on small, non-representative, sometimes decades-apart samples, ad hoc adjustments, and inconsistent test types — problems serious enough that many psychologists regard the dataset as unfit for the cross-national comparisons it’s used for. If you came in wanting “the average IQ” and bumped into a world map of IQ scores: those are different claims, and the map is the shaky one.
Quick reference: where common scores land
Not requirements — just where the population sits, for context. (The good IQ score article has the research-backed version of this with sources.)
| Score | Roughly… |
|---|---|
| ~100 | The middle of the adult distribution; the modal outcome. |
| ~110–115 | Around the average for adults who complete a demanding university degree. |
| ~120 | ~91st percentile; the upper end of “superior.” |
| 130 | ~98th percentile; the standard Mensa-eligibility cutoff. |
| 145+ | ~1 in 1,000; “highly gifted” territory. |
Frequently asked questions
What is the average IQ for an adult?
It’s 100, by definition — IQ scales are built so the mean of the reference sample is 100, with a standard deviation of 15. About two-thirds of adults score between 85 and 115; about 95% score between 70 and 130.
Is 115 a good IQ for an adult?
Yes — it’s one standard deviation above the mean, roughly the 84th percentile, ahead of about five in six adults. It’s in the “high average” to “superior” range and is around the level often associated with completing a demanding degree, though no score is a hard requirement for anything.
What is a low IQ for an adult?
Scores below about 85 fall in the “low average” range (roughly the bottom 16%) and below about 70 in the “extremely low” range (roughly the bottom 2%). As always, the ±5-point margin of error and the testing conditions matter — one score is one data point, not a diagnosis.
Does the average adult IQ change with age?
No — it’s age-normed, so the average is 100 in every adult age band. What changes is the mix behind it: a younger adult’s average leans on processing speed and novel reasoning, an older adult’s on vocabulary and accumulated knowledge. Feeling slower at 50 isn’t the same as a drop in IQ.
Why did an online test say my IQ is over 120?
Most free online tests aren’t standardized on a representative sample, are taken by an above-average self-selected crowd, and have a commercial incentive to score generously — so they tend to run high. Treat a flattering online result with skepticism; a clinically administered test like the WAIS-IV is the only kind that yields a defensible number.
Has the average adult IQ gone up over time?
The raw performance behind it has — that’s the Flynn effect, roughly 3 points per decade through the 20th century — but tests are re-normed so the score stays centered on 100. It means an old IQ and a modern IQ aren’t directly comparable point for point. In several wealthy countries the gains have stalled or slightly reversed since the 1990s.
Is “average national IQ” the same as the average adult IQ?
No. “The average adult IQ is 100” is about how one well-normed test is scaled. The country-ranking figures come from a separate, widely criticised body of work built on small, non-representative samples and inconsistent data, and shouldn’t be treated as a reliable comparison between populations.
Related reading
- Average IQ by Age — the full lifespan picture: why every age band averages 100 and what fluid vs. crystallized intelligence does from childhood to late life.
- Average IQ for a Teenager — the teen years specifically: why there’s no jump at 13, why a bright teen’s grades can slip, and the WISC-to-WAIS handoff at 16.
- Average IQ for a 12-Year-Old — the same question for a child, and why a school report needs reading differently from an adult result.
- What Is a Good IQ Score? — “good” for what: research-backed thresholds for college, careers, and Mensa, plus the margin-of-error reality check.
- IQ Classification Ranges — what each band means and why the boundaries are interpretive, not biological.
- What Your IQ Score Actually Means — the 100-point scale, what it predicts, and what it doesn’t.
References
- Wechsler, D. (2008). Wechsler Adult Intelligence Scale — Fourth Edition (WAIS-IV): Technical and Interpretive Manual. Pearson.
- Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101(2), 171–191.
- 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.
- Cattell, R. B. (1963). Theory of fluid and crystallized intelligence: A critical experiment. Journal of Educational Psychology, 54(1), 1–22.
- Salthouse, T. A. (2009). When does age-related cognitive decline begin? Neurobiology of Aging, 30(4), 507–514.
- Sear, R. (2022). Demography and the rise, apparent fall, and resurrection of eugenics. Population Studies, 76(3), 363–380. (On the methodological problems with “national IQ” datasets.)