What Is the Average IQ — and What Your Score Really Means

The number 100 isn’t a fact about brains — it’s a decision about where to put the middle. Once you see that, every IQ score finally makes sense.

Someone tells you their IQ is 122 and you nod, vaguely impressed, with no real idea what that means. Is 122 a lot? A little? Two points from genius? The truth is that a raw IQ number tells you almost nothing on its own — it only becomes meaningful once you know the shape of the distribution it came from. And that shape is far simpler, and far more deliberately engineered, than most people realize.

Here’s the whole article in one sentence, because the rest is just unpacking it:

An IQ score isn’t a measurement of how much intelligence you “have” — it’s a statement of where you rank in a population that was deliberately arranged so the average is exactly 100.

That single design choice — fixing the average at 100 — is the key that unlocks percentiles, score bands, the bell curve, and the difference between “above average” and “rare.” Get it, and you’ll read any IQ result correctly for the rest of your life.

(If you’d rather just measure where you land, here’s a properly normed test — but the context below is genuinely worth reading first, because a number with no distribution behind it is just a number.)

The average IQ is 100 — and that’s by construction, not by discovery

Start with the fact everyone half-knows: the average IQ is 100.

What almost nobody is told is why. It’s tempting to imagine that researchers measured a huge crowd, added up everyone’s intelligence, divided, and the answer happened to come out at 100. That’s not what happened. 100 is set, not found. It’s defined into existence.

Modern IQ scores are what psychologists call a deviation IQ. When a test like the Wechsler Adult Intelligence Scale is built, the makers give it to a large, carefully selected reference group — the norm sample — that’s matched to the national population by age, sex, region, and education. Then they do something that sounds almost like cheating but is actually the whole point: they take the average raw performance of that group and declare it to be 100. Everyone else’s score is then expressed as a distance from that anchor.

So when your result says 113, it doesn’t mean you possess “113 units” of anything. It means: relative to the norm group, you landed a specific distance above their middle. The 100 is a ruler’s zero-point that someone chose to paint there.

Why 100 specifically? History. The earliest tests used a literal quotient — mental age divided by chronological age, times 100 — so an average child (mental age = actual age) scored exactly 100. When David Wechsler abandoned that flawed ratio method in 1939, he kept the familiar 100 anchor so the new scores would feel continuous with the old, even though the math underneath had completely changed. The number stuck. It’s a historical convention dressed up as a constant of nature.

The bell curve: why most people cluster near the middle

Once the average is pinned at 100, the next question is: how spread out is everyone else? That’s where the famous bell curve comes in.

IQ scores are designed to follow a normal distribution — the symmetric, bell-shaped curve that shows up everywhere from human height to measurement error. It has two defining features, and only two:

• A center — the mean, here fixed at 100.
• A spread — the standard deviation (SD), which on the Wechsler and modern Stanford–Binet scales is 15 points.

The standard deviation is the unsung hero of the whole system. It’s the unit that converts “a number” into “a rank.” One standard deviation up from the mean is 115 (100 + 15). One down is 85 (100 − 15). Two up is 130; two down is 70. The bell curve is tall in the middle and thin at the edges, which is just a visual way of saying the obvious: scores near 100 are common, and scores far from 100 are rare.

This is why “average” in IQ doesn’t mean a single value of 100 — it means a band around 100 where most of humanity sits. The next section makes that precise.

The 68–95–99.7 rule, in IQ terms

Every normal distribution obeys a tidy regularity called the empirical rule, or the 68–95–99.7 rule. Plug in an IQ mean of 100 and an SD of 15, and it translates directly into the numbers you’ve probably half-heard:

• About 68% of people score within one SD of the mean — IQ 85 to 115. Roughly two out of every three people you meet fall in this band. This is what “average intelligence” actually refers to: not the single point 100, but this broad central swath.
• About 95% score within two SDs — IQ 70 to 130. Nineteen people in twenty live inside this range. Step outside it in either direction and you’re in the outer 5%.
• About 99.7% score within three SDs — IQ 55 to 145. Only about 3 people in 1,000 land beyond these walls.

Notice what this does to the word “rare.” An IQ of 115 sounds modest, but it already beats roughly 84% of people. An IQ of 130 — the common cutoff for “very superior” — is hit by only about 1 person in 44. The curve is steep: each additional 15 points doesn’t add a fixed chunk of people, it slices off a progressively smaller and smaller sliver of the population. That non-linearity is why the gap between 130 and 145 feels so much larger than the gap between 100 and 115, even though both are 15 points wide.

Percentiles: the honest way to read your score

If the three-digit IQ number is the jargon, the percentile is the plain-English translation — and it’s the single most useful way to read any result.

Your percentile is simply the share of people you scored at or above. The mapping falls straight out of the bell curve:

• IQ 70 → 2nd percentile — beats about 2% of people
• IQ 85 → 16th percentile — beats about 16%
• IQ 100 → 50th percentile — beats about 50% (the exact middle)
• IQ 115 → 84th percentile — beats about 84%
• IQ 130 → 98th percentile — beats about 98%
• IQ 145 → 99.9th percentile — beats about 99.9%

This is why 100 is, by definition, the 50th percentile — exactly half score above it and half below. It’s the literal middle of the line.

Percentiles also quietly puncture some bragging. A 30-point gap between two friends — say 100 vs. 130 — looks huge as a number, and it is: it’s the difference between the median and the top 2%. But a 30-point gap between 130 and 160 represents far fewer actual people, because there’s almost no one left out there to separate. Percentiles keep you honest about how crowded — or empty — your part of the curve really is. When you eventually go and measure your own score, the percentile is the figure to actually trust.

The score-band labels (and a warning about taking them too seriously)

Test publishers attach verbal labels to ranges of scores so that a clinician can describe a result in words. The standard Wechsler classification looks roughly like this:

• 130 and above — Very Superior / Extremely High (98th percentile and up)
• 120–129 — Superior (91st–98th)
• 110–119 — High Average (75th–90th)
• 90–109 — Average (25th–75th)
• 80–89 — Low Average (9th–25th)
• 70–79 — Borderline (2nd–9th)
• 69 and below — Extremely Low (below 2nd)

Two things are worth flagging. First, the “Average” label covers 90–109, not just 100 — a span that, by itself, contains roughly half the population. So “you’re average” is statistically true of an enormous number of people and shouldn’t sting. Second, these labels are conventions, and they’ve drifted over the decades to use gentler language (older charts used now-retired clinical terms). The labels are a communication tool, not a law of nature — the percentile underneath is the real information.

Two footnotes that quietly change the number: SD 16 and the Flynn effect

Two technical wrinkles trip up a lot of online IQ chatter, and both are worth knowing.

The SD-16 variant. Not every test uses a standard deviation of 15. The older Stanford–Binet (Form L-M) and a few other instruments used SD 16, and the Cattell scale famously uses SD 24. This matters because the same person at the same true ability gets a different number depending on the scale. Being two SDs above the mean is an IQ of 130 on a Wechsler (SD 15) but 132 on the old Stanford–Binet (SD 16) — which is exactly why Mensa lists its “top 2%” cutoff as 130 on one scale and 132 on another. They describe the same rarity. So a bare “my IQ is 132” is ambiguous until you know which ruler produced it. This is also a fast way to spot dodgy online tests: a transparent one tells you its SD; a number-generator never does.

The Flynn effect. Here’s the strangest fact in the whole field. Raw cognitive-test performance rose steadily across the 20th century — on the order of about three IQ points per decade, a phenomenon named after researcher James Flynn. If the average were left to float freely, it would have drifted well above 100 over time. It doesn’t — because test publishers periodically re-norm their tests, re-collecting a fresh sample and resetting that sample’s average back to 100. That’s the deeper reason 100 is a moving anchor that’s held fixed by design: every generation is re-centered against itself. Score a person from 1932 against today’s norms and they’d look far below average; score today’s population against 1932 norms and they’d look like geniuses. The number 100 is stable precisely because the people defining it are constantly being refreshed.

The bottom line

An IQ score is a rank, not a quantity. It tells you where you sit in a population that was deliberately arranged so the average is 100 and the spread is 15. Everything else follows mechanically from those two numbers:

• 100 is the 50th percentile — the engineered middle, not a measurement of “enough” intelligence.
• 85–115 is the average band holding ~68% of people; 70–130 holds ~95%.
• Percentiles are the honest translation of the three-digit number, and they reveal how steeply rarity climbs as you move away from the center.
• The scale matters (SD 15 vs. 16), and the anchor is maintained, not discovered, thanks to the Flynn effect and re-norming.

Read your own score through that lens and it stops being a verdict on your worth and becomes what it actually is: a coordinate on a well-understood curve.

FAQ

Is 100 a “good” IQ? It’s the exact average — the 50th percentile — so it’s neither good nor bad; it’s the definition of typical. Roughly half of all people score below 100 and half above. The whole 90–109 range is officially labeled “Average,” and it contains about half the population.

Why is the average always exactly 100 instead of some messier number? Because it’s set there by design, not measured. Test makers define the norm group’s average performance as 100, then express everyone else as a distance from it. When raw performance drifts over the years (the Flynn effect), the test is re-normed to pull the average back to 100.

What’s the difference between IQ 130 and IQ 132? Usually nothing about the person — it’s the scale. 130 is two standard deviations above the mean on a SD-15 test (Wechsler); 132 is two SDs above on a SD-16 test (older Stanford–Binet). Both describe the same top-2% rarity. Always ask which standard deviation a score is based on.

How rare is a “genius-level” IQ? There’s no official “genius” cutoff, but using common thresholds: about 1 person in 44 reaches 130 (98th percentile), and only about 1 in 1,000 reaches roughly 145 (three SDs up). The curve thins dramatically, so each extra 15 points removes a far larger share of the remaining people.

My online test gave me a much higher number than I expected — is it real? Possibly, but be skeptical of any test that doesn’t state its standard deviation, its norm sample, or how it maps scores to percentiles — those are the things that make 100 mean anything. A flattering number with no distribution behind it is decoration, not measurement. (The companion piece on whether online IQ tests are accurate has a quick checklist for telling the two apart.)

If you’d like to see where you actually land, you can try a test built on these principles — transparent norms, a sensible length, percentile and score band shown plainly, and no paywall ambush at the end.

[embed]IQ Test 2026 | Accurate Intelligence Diagnosis Discover your true intelligence in just 5 minutes. Uncover your potential with accurate and modern IQ diagnosis based…iq-test-official.site