What is a good IQ score?

The short answer: 100 is average by definition, about two thirds of people score between 85 and 115, and anything above 115 puts you in roughly the top sixth of the population. The longer answer is more interesting, because "good" depends entirely on understanding what the scale actually measures and how it is built.

How the IQ scale is constructed

IQ is not measured the way height or weight is measured. There is no natural unit of intelligence, so the scale is built statistically. A test is given to a large, representative sample of people, and the distribution of raw scores is mapped onto a normal distribution — the bell curve — with the average fixed at 100 and the standard deviation fixed at 15 points.

That construction has three consequences worth understanding before judging any score as good or bad. First, 100 is always the middle: half of people score above it and half below, in every generation, because the scale is periodically re-anchored to keep it that way. Second, the further a score sits from 100, the rarer it is — and rarity accelerates fast. Moving from 100 to 115 takes you past about a third of the population; moving from 115 to 130 takes you past most of the rest. Third, small differences near the middle of the scale are common and largely meaningless: an IQ of 102 and an IQ of 98 describe people whose measured reasoning ability is, for practical purposes, indistinguishable.

The standard classification bands

Psychologists group scores into descriptive bands. Labels vary slightly between test publishers — the convention below follows the Wechsler scales published by Pearson Clinical, the most widely administered adult tests — with the share of the population falling in each band.

Notice how much of the population the average band swallows: half of everyone you have ever met scores between 90 and 109. If your score lands there, you are in the company of teachers, tradespeople, managers, nurses, and most of the people running the world day to day. The average band is not a consolation prize — it is the human norm the whole scale is anchored to.

So what actually counts as good?

It depends on the comparison you care about. Statistically, any score above 100 beats the population average, and a score of 110 already places you ahead of roughly three quarters of people. A score of 115 — one standard deviation above the mean — is the level many people intuitively mean by "smart": comfortably top sixth, common among university graduates, and typically well above the demands of most jobs.

Scores of 120 to 129 are where things become genuinely uncommon: fewer than one person in ten reaches this band, and it is heavily represented in cognitively demanding professions such as medicine, law, engineering, and research. At 130 and above you are past ninety-eight people in a hundred, which is the threshold most high-IQ societies use for membership.

But here is the part most articles skip: the practical value of additional IQ points is not linear. The difference between 85 and 100 tends to matter a great deal in daily life — it shows up in how easily someone handles paperwork, plans, instructions, and new systems. The difference between 115 and 130 matters in narrower, more specialised contexts. And beyond roughly 120, research on real-world outcomes suggests diminishing returns: conscientiousness, opportunity, and social skill increasingly dominate. A good IQ score opens doors; it does not walk you through them.

Percentiles: the number that matters more than the score

The most useful way to read any IQ score is as a percentile — the percentage of people you scored as high as or higher than. The score itself is an arbitrary rescaling; the percentile is the plain-English fact underneath it.

• IQ 100 is the 50th percentile: higher than half of people.
• IQ 110 is roughly the 75th percentile: higher than three in four.
• IQ 115 is roughly the 84th percentile: higher than five in six.
• IQ 120 is roughly the 91st percentile: higher than nine in ten.
• IQ 130 is roughly the 98th percentile: higher than forty-nine in fifty.

Percentiles also expose why chasing tiny score differences is pointless. Between 100 and 105 lies about 13% of the population — a crowd. Between 140 and 145, less than a fifth of one percent. The same five points mean completely different things depending on where on the curve they fall.

What a single test score can and cannot tell you

Every IQ score is an estimate with measurement error, not a reading from a dial. The same person tested twice will typically score within a few points of themselves, but not identically — sleep, stress, motivation, familiarity with the question format, and plain luck on the day all move the needle. Professionally administered assessments report a confidence interval around the score for exactly this reason, and any serious interpretation of an online test result should carry the same humility.

A score also compresses many distinct abilities into one number. Two people with identical overall scores can have very different cognitive profiles — one stronger in verbal reasoning, another in spatial or numerical pattern work. Matrix reasoning tests such as ours deliberately target the pattern-and-logic core that correlates most strongly with general ability, which makes them fair across languages and backgrounds, but no single format captures everything a mind does.

Does a good score matter in real life?

At the level of large groups, IQ is one of the better predictors psychology has: it correlates with school performance, speed of learning new job skills, and performance in complex work. That is precisely why armed forces, some employers, and selective schools have used cognitive testing for a century.

At the level of one individual — you — the honest answer is more modest. Plenty of people with unremarkable scores build remarkable lives, and plenty of high scorers do not. The traits that consistently compound over a lifetime are showing up, following through, staying curious, and getting along with people. A good IQ score is a genuine advantage, roughly like being tall in basketball: helpful, real, and nowhere near the whole game.

Can you raise your IQ?

Mostly no, a little yes, and the details matter. Underlying general ability is substantially stable in adulthood: retest gains of a few points come from familiarity with question formats, not from becoming more intelligent, and commercial brain-training products have repeatedly failed to show transfer beyond the trained games in controlled studies. The honest levers are slower and less marketable. Additional years of education are associated with modest score gains. Aerobic exercise, sleep, and treating hearing and cardiovascular problems protect the machinery, especially with age. And on any specific test, practice with the format genuinely helps performance — which is exactly why comparing your score only with people taking the same test, as this site does, is the fair comparison.

Same score, different labels

Publishers attach slightly different words to the same numbers, which causes endless confusion. The bands are anchored to the same statistics; only the vocabulary shifts.

When you see a dramatic-sounding label attached to a score, translate it back to the percentile before reacting: the words are marketing-adjacent, the percentile is the maths.

How scores are actually used today

Understanding where testing is used — and where it is not — helps calibrate how much any score should matter to you. Formal cognitive testing survives in a handful of places: educational psychology, where supervised assessments feed decisions about learning support and gifted programmes; clinical settings, where profiles across sub-tests matter more than the headline number; some militaries, which have screened recruits with cognitive batteries for a century; and parts of graduate recruitment, usually as aptitude tests that are IQ tests in all but name. Outside those settings, hardly anyone will ever ask for your IQ. No interviewer requests it, no university admits on it, and volunteering it is generally read as odd. The number's real use is private: it is decent information about how easily you are likely to absorb complex, novel material, which can sensibly inform how you study, what work energises you, and how much of a challenge to bite off — and that is roughly the limit of its jurisdiction over your life.

It is also worth knowing that employment testing is legally constrained in many jurisdictions. Employers who screen on cognitive tests must generally show the test relates to the job's actual demands, precisely because blanket IQ gates have historically produced discriminatory outcomes. The practical upshot for you as an individual: no single score will ever formally open or close a door — the doors that matter are opened by demonstrated skills, and scores only ever whisper in the background.

Frequently asked questions