Mensa gets all the headlines, but it's actually the least exclusive of the well-known high-IQ societies. High-IQ societies are membership organisations that admit people based on a qualifying score on a recognised IQ (intelligence quotient, a score that ranks reasoning ability against the general population) test, and several set the bar far higher than Mensa's top 2%. Intertel asks for roughly the top 1%. The Triple Nine Society wants the top 0.1%. A handful of even smaller groups sit above that. This piece walks through what those thresholds actually mean, in real numbers of people, and why "more exclusive" isn't automatically "more meaningful."
What counts as a high-IQ society?
A high-IQ society is, at its core, a club with one unusual entry requirement: a test score. Instead of vetting members by profession, wealth, or interest, these groups use a percentile — a statistic that tells you what share of the population a given score outranks — as the gatekeeper. Score above the group's cutoff on an accepted test, submit the proof, pay a membership fee, and you're in.
Mensa is by far the largest and oldest, founded in England in 1946 and now present in dozens of countries. If you want the full picture of how it works, what Mensa actually is and Mensa's qualifying scores across different tests are both covered in detail elsewhere on this site. But Mensa's 98th-percentile bar — meaning a person scores higher than 98 out of 100 people — is only the entry point into a much wider ecosystem of societies, each pitched at a rarer slice of the bell curve.
Why percentile matters more than the raw IQ number
Here's the thing that trips people up: different tests report scores on different scales, so a "132" on one test and a "148" on another can represent the same rarity. That's why serious high-IQ societies define their requirement by percentile, not by a fixed IQ number. The percentile is the constant; the number attached to it depends on which test, and which scale, produced it.
This matters because standard IQ scales are built around a mean of 100 and a standard deviation of 15 (a measure of how spread out scores are around that average). On that curve, a given percentile always corresponds to roughly the same distance from the mean — but a test using a standard deviation of 16 or 24 will attach a different-looking number to that same spot. If this is new territory, the IQ bell curve explainer and the IQ percentile guide both go through the maths without the jargon.
Mensa vs. Intertel vs. Triple Nine: the percentile ladder
The easiest way to see how these societies stack up is to line up their thresholds and translate each one into "people per thousand" — a more intuitive way to grasp rarity than a percentile figure alone.
| Society | Percentile threshold | Roughly how many people per 1,000 qualify | Approx. IQ (SD 15 scale) |
|---|---|---|---|
| Mensa | 98th | 20 per 1,000 | ~130 |
| Intertel | 99th | 10 per 1,000 | ~135 |
| Triple Nine Society | 99.9th | 1 per 1,000 | ~146 |
| Prometheus Society (for reference) | 99.997th | ~3 per 100,000 | ~160 |
A few honest caveats belong right next to that table. First, the IQ-equivalent column is approximate — it depends heavily on which test produced the underlying score, and at the extreme tail, scores from different tests diverge more, not less. Second, very high percentiles (99.9th and beyond) are statistically harder to measure precisely, because so few people in any test's norming sample actually scored that high — the tail of the bell curve gets thin fast. Societies at that level often accept a narrower list of tests specifically because of this measurement problem.
What "1 in 1,000" actually feels like
It's easy to read "99.9th percentile" and let the number wash past you. Translated literally: in a stadium of 10,000 people, roughly 200 would qualify for Mensa, 100 for Intertel, and only about 10 for the Triple Nine Society. That's the practical difference between "top 2%" and "top 0.1%" — not a subtle shift, but an order-of-magnitude drop in rarity.
Intertel: the "top 1%" alternative to Mensa
Intertel was founded in the United States in 1966 and requires a score at or above the 99th percentile on an approved test. It's smaller and lower-profile than Mensa, with less emphasis on local social events and more on newsletters, correspondence, and a member directory of specialists across many fields. Functionally, it sits as a middle step between Mensa's broad, social 98th-percentile club and the far narrower societies above it.
The Triple Nine Society: three nines, one clue in the name
The Triple Nine Society (often shortened to TNS) takes its name directly from its requirement: the 99.9th percentile — three nines in a row. Founded in 1978, it accepts a specific list of qualifying tests and scores, similar in spirit to how Mensa vets applicants, but the accepted cutoff on each test sits much further out on the curve. Because 99.9th-percentile scores are rare and harder to pin down reliably, TNS tends to be stricter about which tests and which exact score reports it will accept.
Beyond Triple Nine: the ultra-rare societies
A small number of societies go even further, some claiming thresholds around the 99.99th percentile or beyond — figures that put membership somewhere in the single digits per 100,000 people. These groups are tiny by design, often numbering in the dozens or low hundreds of members worldwide. Two things are worth knowing about this tier:
- Measurement gets shakier the further out you go. Test norms are built from real samples of real test-takers, and almost no test is normed on enough people to reliably distinguish, say, the 99.99th percentile from the 99.97th. Claims at this level deserve a healthy amount of scepticism.
- Verification methods vary a lot. Some ultra-high societies accept a narrow set of specific, heavily documented tests; others have looser or more informal review processes. That inconsistency is part of why mainstream psychology treats these extreme percentile claims cautiously.
Are stricter societies "better," or just narrower?
Neither, really — they're different filters for different goals. A stricter percentile cutoff mathematically produces a smaller, more homogeneous group by score, but it says nothing about warmth, community activity, or what members actually do together. Mensa's broader threshold means a bigger, more socially active membership with local chapters and events in many cities; a group like Triple Nine trades that scale for rarity and a more specialised, low-key correspondence-style community.
It's also worth separating the statistical exclusivity from the personal payoff. If you're weighing whether a society membership is worth pursuing at all, this honest look at whether Mensa is worth it raises questions — networking value, sense of belonging, practical perks — that apply just as much to Intertel or Triple Nine as they do to Mensa. A higher cutoff doesn't automatically mean a better fit for you.
A word on what a score does and doesn't tell you
None of these thresholds measure character, creativity, emotional intelligence, or how well someone applies their reasoning ability day to day. They measure performance on a specific, standardised reasoning test, on one occasion, compared against a norming sample. That's a real and useful thing to know about yourself — but it's one data point, not a verdict on your worth or potential. If you want the fuller picture of what a "high IQ" traditionally implies, and what it doesn't, this breakdown of high-IQ traits and myths is a useful next read.
Key takeaways
- High-IQ societies rank by percentile, not a fixed IQ number, because different tests use different scales for the same rarity.
- Mensa (98th percentile) is the largest and least exclusive of the well-known societies; Intertel (99th) and the Triple Nine Society (99.9th) are progressively rarer.
- Translating percentiles into "people per 1,000" makes the real rarity gap between these thresholds much easier to grasp than the raw percentage.
- Claims above the 99.9th percentile become statistically shakier, since almost no test is normed on a large enough sample to measure that tail reliably.
- A stricter cutoff means a smaller, more score-homogeneous group — not automatically a better community fit or a more meaningful achievement.
None of this replaces a proper qualifying test if you're pursuing a real society application. iqmetria's free test is orientative — a fast, genuinely useful way to get a sense of your reasoning ability for your own curiosity, not an official or clinical assessment, and not affiliated with Mensa, Intertel, or any high-IQ society.
FAQ
What IQ do you need for Intertel?+
Intertel requires a score at or above the 99th percentile on an approved test — roughly the top 1% of the population, which is about 10 people out of every 1,000. On a standard SD-15 IQ scale, that lands around IQ 135, though the exact number depends on which test was used.
What percentile is the Triple Nine Society?+
The Triple Nine Society requires the 99.9th percentile, meaning only about 1 person in 1,000 qualifies. That's roughly ten times rarer than Mensa's 98th-percentile requirement.
Is there anything more exclusive than the Triple Nine Society?+
Yes — a small number of ultra-selective societies claim thresholds around the 99.99th percentile or beyond, with only a handful of qualifying people per 100,000. Claims at this level are harder to verify reliably, since very few tests are normed on samples large enough to measure that extreme tail with confidence.
Are high-IQ societies like Mensa and Intertel worth joining?+
It depends on what you're after — community, networking, or simply confirming a score, rather than any inherent 'better' or 'worse' between societies. A stricter percentile cutoff makes a group rarer, not necessarily more valuable to you personally.
Do all high-IQ societies require the same test?+
No. Each society keeps its own list of approved, standardised tests and sets a specific qualifying score or percentile for each one. Rarer societies (like Triple Nine) tend to accept a narrower, more strictly documented list because ultra-high scores are harder to measure precisely.