Mensa and high-IQ societies: what it takes to qualify
Mensa asks for the top 2%. That single criterion produces several different qualifying numbers depending on the test, which is the source of nearly all the confusion about it.
On this page
One criterion, several numbers
Mensa was founded in 1946 with a deliberately simple entry rule that has not changed: a score at or above the 98th percentile on an approved, properly administered intelligence test. That is the whole criterion. Everything else is implementation.
The confusion starts because "98th percentile" converts to different IQ numbers on different scales. IQ scales all set the mean at 100, but they do not all use the same standard deviation.
| Test | Standard deviation | 98th percentile score |
|---|---|---|
| Wechsler scales (WAIS, WISC) | 151 | ~130 |
| Stanford-Binet (older forms) | 163 | ~132 |
| Cattell III B (used by British Mensa) | 24 | ~148 |
This is why someone saying "my IQ is 148" may be describing exactly the same ability as someone saying "my IQ is 130". An IQ number is meaningless without knowing the scale it came from. The percentile is the portable statement; the IQ figure is not.
The 98th percentile means roughly 1 person in 50. That is selective, but it is not rare in the way the surrounding mystique implies - in a large secondary school there will be several dozen people who would qualify.
The two routes in
1. Sit Mensa’s own supervised admission test. Administered under proctored conditions by the national chapter. The test used varies by country; some chapters administer a battery of two instruments where reaching the threshold on either is sufficient.
2. Submit prior evidence. A qualifying score from an approved test already taken under proper conditions - typically a clinical instrument such as a Wechsler scale or Stanford-Binet administered by a qualified psychologist. Some chapters also accept certain standardised admissions tests at specified thresholds.
Which instruments qualify, and at what cut, varies by national chapter and changes over time. Check your own country’s Mensa organisation rather than relying on any third-party list, including this one.
Why online scores never count
No unsupervised online test can qualify you for Mensa, and this is not gatekeeping for its own sake. There are four solid reasons, and they apply to this test as much as any other.
- Identity and conditions are unverified. Nobody can confirm who sat the test, whether they were interrupted, or whether they looked anything up.
- Item exposure. A fixed public item set leaks. Once answers circulate, scores stop measuring reasoning and start measuring exposure.
- Norms are usually absent. A defensible score requires a standardisation sample matched to a target population. Most free tests have none - and say nothing about it.2
- Precision collapses at the tails. This is the decisive one for a 98th-percentile decision.
That last point deserves expanding, because it explains a design decision on this site. A test estimates ability most precisely where it has the most items near your level. A 30-item test has few items hard enough to discriminate among the top 2%, so the confidence interval widens sharply out there. Reporting "IQ 132" from a short test implies a precision the test cannot deliver.
This is exactly why results here are capped at 65 and 135 and reported as an interval rather than a point. A score of "135 or above" is an honest statement about what a 30-item test can resolve. "IQ 148, Mensa level" would not be.5
The other societies
Mensa is the largest and least restrictive. Several smaller societies set higher cuts:
| Society | Percentile | Approx. rarity |
|---|---|---|
| Mensa | 98th | 1 in 50 |
| Intertel | 99th | 1 in 100 |
| Triple Nine Society | 99.9th | 1 in 1,000 |
| Prometheus Society | 99.997th | ~1 in 30,000 |
A statistical caution about the bottom of that table. Distinguishing the 99.9th percentile from the 99.997th requires a test with enough sufficiently difficult items and a standardisation sample large enough to contain meaningful numbers of people that rare. A sample of 2,200 adults - typical for a major clinical test - contains perhaps two people above the 99.9th percentile. Norms that far out are extrapolation from a fitted curve, not observation.
Add the Flynn effect on top, which shifts raw performance by roughly 0.28 points a year and makes the vintage of the norms material,4 and scores quoted above about 145 should be treated with real caution regardless of source.
What membership does and does not mean
Membership means one thing precisely: on one supervised test, on one day, you scored in the top 2% of a reference group. It is a legitimate, verifiable fact.
It does not mean you will do well at work, be right in arguments, or be good at anything in particular. The correlations between cognitive ability and real outcomes are real but loose - around 0.31 with job performance and 0.23 with income - as set out in what IQ actually predicts. A threshold on one measure carries no more weight than that measure does.
The honest case for joining is social rather than evidential: people who enjoy this kind of puzzle often enjoy the company of others who do. That is a perfectly good reason. Treating the membership card as a verdict on your worth is not.
If you want to try: take this test for interest by all means, but for a qualifying score contact your national Mensa organisation and sit their supervised test. Nothing you do unsupervised at home can substitute for it.
Common questions
What IQ do you need for Mensa?
The 98th percentile on an approved supervised test - about 130 on a Wechsler scale (SD 15), 132 on older Stanford-Binet forms (SD 16), or about 148 on the Cattell III B (SD 24) used by British Mensa. All three describe the same level of ability on different scales.
Why do people quote different Mensa IQ numbers?
Because IQ scales use different standard deviations. The percentile is the meaningful statement; the IQ number depends entirely on which scale produced it. An IQ figure quoted without its scale is not interpretable.
Can an online IQ test qualify me for Mensa?
No. Mensa accepts only supervised tests taken under verified conditions, either its own admission test or an approved instrument administered by a qualified professional. That includes this test - no unsupervised result can establish eligibility.
Why does this test not report scores above 135?
Because it cannot resolve them honestly. A 30-item test has too few very hard items to distinguish reliably among the top few percent, so the confidence interval widens sharply at the extremes. Reporting a precise high number would imply a precision the test does not have.
References
- Wechsler, D. (2008). Wechsler Adult Intelligence Scale - Fourth Edition: Technical and Interpretive Manual. Pearson.
- American Educational Research Association, American Psychological Association, & National Council on Measurement in Education (2014). Standards for Educational and Psychological Testing. AERA.
- Roid, G. H. (2003). Stanford-Binet Intelligence Scales, Fifth Edition: Technical Manual. Riverside Publishing.
- 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. doi:10.1177/1745691615577701
- Condon, D. M., & Revelle, W. (2014). The International Cognitive Ability Resource: Development and initial validation of a public-domain measure. Intelligence, 43, 52-64. doi:10.1016/j.intell.2014.01.004
Related reading
Find out where you land
Thirty questions, about 30 minutes, and a score with the uncertainty attached.
Start the test30 questions · about 30 minutes · no sign-up