Mensa admits people who score in the top 2% of the general population on an approved, supervised intelligence test. There is no single qualifying number, because the cutoff depends on which test you take: 132 on the Stanford-Binet, 148 on the Cattell III B, and 130 on the Wechsler scales. All three describe the same 2% threshold expressed on different scales.

Why one society has several cutoffs

The confusion comes from standard deviation. IQ scales set the average at 100, but they differ in how many points represent one standard deviation from it.

TestStandard deviationTop 2% cutoff
Wechsler (WAIS, WISC)15130
Stanford-Binet16132
Cattell III B24148

The percentile is the real requirement and the score is just its translation into whatever scale the test uses. Someone with a Cattell score of 148 has not outperformed someone with a Wechsler 130. They took a test that spreads its scores more widely.

This is also why a score quoted without its test is close to meaningless. A claimed IQ of 148 could be the 99.98th percentile or the 98th, depending entirely on which instrument produced it.

Which tests Mensa accepts

American Mensa accepts more than 200 qualifying tests, and the list reaches well beyond instruments with IQ in the name. Older SAT and GRE results, the LSAT, the CogAT, the Miller Analogies Test, and various school-administered batteries all appear on it, subject to date cutoffs, since scoring changes over time invalidate older versions.

Two conditions apply to all of them. The test must have been properly administered by a qualified examiner, and it must have been supervised. A score you produced alone at home does not qualify no matter which test it came from.

The other route is Mensa's own supervised admission test, taken at a scheduled session. It is the simplest option for most people because it removes the question of whether prior scores are eligible.

Online tests do not qualify, including this one

No unsupervised online test can be used for Mensa admission. That covers every free IQ test on the internet and it covers ours. The reason is not quality; it is that nobody can verify who took the test, under what conditions, or with what assistance.

Where an online test is genuinely useful is as a filter. If you consistently score around 110 on well-constructed practice tests, the odds of clearing a supervised 98th-percentile threshold are low, and the sitting fee is real money. If you are regularly landing at 130 and above, booking the supervised test is a reasonable bet.

Treat that as a rough signal rather than a prediction. Online scores are noisier than supervised ones, and the direction of the error is usually generous.

Is membership worth pursuing

Mensa is a social organisation. It runs local groups, national gatherings, special interest groups, and publications. What it does not do is confer any professional credential, and no employer treats membership as a qualification.

The honest case for joining is that the local group appeals to you. The honest case against is that a high test score selects for one narrow thing, and a room full of people selected on it is not automatically a room you want to spend evenings in.

It is worth knowing that the 2% threshold is not a meaningful boundary in nature. Cognitive ability is continuous, and there is no discontinuity at the 98th percentile. The cutoff is an administrative decision about how large the society should be.

General IQ Test

25 questions, 31 minutes, scored as soon as you finish.

Take the General IQ Test

Questions people also ask

Can I join Mensa with an old SAT score?

Sometimes. American Mensa accepts SAT scores from before specific rescoring dates, and the qualifying thresholds differ by era. Scores from recent administrations are generally not accepted. Check the current qualifying-test list rather than relying on what was true a few years ago.

What is the highest IQ society?

Several societies set thresholds above Mensa's, with cutoffs at the 99.9th percentile and beyond. At those levels the statistics get shaky: standardised tests are not normed precisely enough at the extremes to distinguish reliably between candidates, so the societies often rely on their own unsupervised tests.

Keep reading

Dr. Sarah Chen is a cognitive psychology researcher focused on intelligence assessment and psychometrics, and writes here about how tests are constructed, scored, and misread. More guides
← All guides See all nine assessments →