The short version: there is no single agreed "smartest person in the world" because IQ tests aren't designed to reliably measure anything above about 160, and most of the huge numbers you see quoted (William James Sidis, Marilyn vos Savant) are decades-old, unverified, or based on tests that have since been discredited for scoring that high. The closest thing to a credible modern record is Kim Ung-yong's Guinness-recognised 210, and even that comes with caveats worth knowing before you repeat it.
The names that always come up
Ask anyone to name the smartest person who ever lived and you get the same handful of names back. Here's what's behind each one.
- William James Sidis - American child prodigy, enrolled at Harvard aged 11 in 1909, reportedly spoke several languages before he was eight. His claimed IQ of 250 to 300 comes from his biographer, not from a documented test score. It has never been verified.
- Marilyn vos Savant - listed in the Guinness Book of Records between 1986 and 1989 with a score of 228, taken on the Stanford-Binet at age 10. That score was a "ratio IQ" (mental age divided by chronological age), a method that mathematically inflates results in children and was already falling out of use by the time she was tested. Guinness retired the "Highest IQ" category altogether in 1990 because the methodology was too shaky to keep defending.
- Kim Ung-yong - South Korean former child prodigy, Guinness-recognised with a score of 210, enrolled at Hanyang University at age three and worked with NASA as a child researcher. He later said the attention did him more harm than good and chose a quieter academic career as an adult.
- Terence Tao - Fields Medal-winning mathematician, commonly cited with a childhood IQ estimate around 220 to 230 (the exact source of that figure is murky), youngest ever gold medallist at the International Mathematical Olympiad at age 13.
- Christopher Langan - American researcher sometimes dubbed "the smartest man in America" by the press, with an estimated IQ between 195 and 210 from unofficial testing.
Notice the pattern: every single one of these numbers comes from childhood testing, outdated methodology, media reporting, or self-reported claims. None of them is a clean, modern, peer-reviewed adult IQ score. That's not a gap in the research. That's the whole point.
How IQ is measured, step by step
Modern IQ testing (Wechsler Adult Intelligence Scale, Stanford-Binet 5) works like this:
- You sit a series of timed subtests covering verbal comprehension, working memory, processing speed, and visual-spatial reasoning.
- Your raw score on each subtest gets converted into a scaled score based on how people your age typically perform.
- Those scaled scores combine into a full-scale score, standardised so the average is always 100 and the standard deviation is 15.
- A score of 130 puts you roughly in the top 2% of the population. A score of 145 puts you in the top 0.1%.
- Above about 160, the test simply runs out of people to compare you to. The norming samples aren't large enough to statistically distinguish a 160 from a 190, so any score above that range is effectively a best guess dressed up as a precise number.
That last point is the one most "highest IQ in the world" listicles skip over entirely, because it ruins the fun of the headline.
The uncomfortable bit about "record" IQs
Here's what most of these articles won't say plainly: a claimed IQ of 200-plus is not a measurement in any meaningful statistical sense. It's a number generated by a test that was never calibrated to distinguish scores that extreme, often administered decades ago, often to a child, often by someone with an interest in the result being impressive. Guinness World Records knew this, which is why they dropped the "Highest IQ" category in 1990 and have never reinstated it. Mensa, the high-IQ society, caps membership verification at scores the test can support, typically around the 148 to 161 range depending on which test you sat.
So when you see a list ranking "the 10 smartest people ever" with numbers like 230 and 250 next to each other, treat it the way you'd treat a horoscope. Entertaining, not evidence.
A worked example: the IQ claim on a CV
Say you run a small recruitment agency and a candidate's CV mentions "IQ of 189, tested privately." What do you do with that?
- Ask which test it was and when. A childhood ratio IQ from a school assessment at age nine tells you almost nothing about how they'll perform in a project manager role at 34.
- Check whether it's verifiable. Legitimate adult testing (administered by a chartered psychologist) comes with documentation. Self-reported numbers from online tests do not.
- Weigh it against actual evidence of the skill you need. A high IQ score correlates loosely with faster pattern recognition and learning speed, but it tells you nothing about reliability, communication, or whether they'll hit deadlines.
- Drop it from the shortlist criteria entirely if you can't verify it, and judge the work sample instead.
That's the same logic that applies to the famous names above. A number without a verifiable, modern, standardised test behind it is a story, not a fact.
Does a high IQ predict success?
Loosely, yes, up to a point. Longitudinal studies (the Terman studies starting in the 1920s being the best known) found correlation between childhood IQ and later academic and career achievement, but the effect flattens well before you get anywhere near the "genius" range, and plenty of the Terman group's highest scorers went on to unremarkable careers. Meanwhile some of the most commercially successful people in history never published a test score at all.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Stephen Hawking was once asked what his IQ was and refused to answer, saying people who boast about their IQ are losers. That's worth sitting with if you've ever used a number like that to decide who's worth listening to in a meeting.
If your actual goal isn't "who's the smartest" but "how do I build something that works," the honest answer is that execution beats raw IQ almost every time. You don't need a 160 to build a one-person AI business, you need consistency, a narrow offer, and the discipline to ship. Plenty of people skip the founder-genius story completely and instead learn how to buy a profitable app that's already working rather than build cleverness from scratch.
Where this fits if you write about it for your own audience
If you're covering this topic for your own blog or newsletter because it's a reliable search draw (it is, people search "smartest person in the world" constantly), don't stop at the listicle. Pull one verified number, one unverified claim, and one practical takeaway, same shape as this post. Then turn it into more than one format. A single researched post like this can become a short video explainer, a LinkedIn carousel on how IQ tests work, and a podcast segment, which is exactly the kind of multiplication covered in our guide to content repurposing tools. One piece of real research, five pieces of content, and you've said something the thinner versions of this article never bothered to check.
Mensa and what score gets you in
If you want a number you can trust, this is it. British Mensa requires a score at or above the 98th percentile on an approved test, which usually means 148 on the Cattell III B scale or 132 on the Wechsler scale. American Mensa uses the same top-2% threshold, roughly 130 on most standard tests. Fewer than one in fifty people qualify. That's a real, checkable bar. Everything above that, into the 180s, 200s, and 250s you see in clickbait lists, is outside the range any current test can reliably confirm.
Frequently asked questions
Who has the highest IQ ever officially recorded?
Kim Ung-yong holds the Guinness-recognised record at 210, set in the 1980s. Guinness retired the "Highest IQ" category in 1990 because testing methods at that extreme end weren't considered reliable enough to keep ranking.
Is Marilyn vos Savant's IQ of 228 real?
It was a genuine test result, but it used a ratio IQ method on a 10-year-old, which mathematically produces inflated scores and is no longer used by modern psychologists. It's accurate to the test given, not comparable to a modern adult score.
What IQ score gets you into Mensa?
Roughly the top 2% of the population, which typically means 132 on the Wechsler scale or 148 on the Cattell III B scale, depending on which test you sit.
Does a high IQ guarantee success in business or life?
No. IQ correlates loosely with faster learning and pattern recognition, but long-term studies show the link to real-world achievement weakens well before you reach "genius" territory. Consistency, communication, and finishing what you start matter far more in practice.
