Ask a room of people whether 91 is prime, and a striking number will say yes without hesitating. It's odd. It doesn't end in 0 or 5. It's not an obvious multiple of anything small. Every instinctive shortcut people reach for says "prime." And every one of them is wrong: 91 = 7 × 13.
Why 91 Fools People Specifically
Most of us check primality with a mental checklist, even if we've never written it down: is it even? Does it end in 5? Is the digit sum divisible by 3? 91 fails all three tests cleanly — it's odd, ends in 1, and its digit sum (9+1=10) isn't divisible by 3. Those three checks cover divisibility by 2, 5, and 3, which between them catch the vast majority of small composite numbers. What they don't cover is 7, and 91 happens to be exactly 7 times 13 — a composite number whose only path to being caught is a divisor most people don't habitually check.
The Actual Math
The reliable way to check primality is trial division by every prime up to the square root of the number. √91 ≈ 9.54, so the complete list of primes you need to test is 2, 3, 5, and 7. 91 fails the first three (as shown above) but 91 ÷ 7 = 13 exactly — a clean, whole-number result. That's the whole story: 91 has a prime factor that just happens to sit past where most people's mental checklist runs out.
A Small Family of Similarly Deceptive Numbers
Once you know to look for it, a whole family of "7-trap" numbers emerges: 217 (=7×31), 259 (=7×37), and 287 (=7×41) all share 91's exact profile — odd, doesn't end in 5, digit sum that clears the divide-by-3 check, and a factor of 7 sitting quietly underneath. A second, smaller family hides behind 11 or 13 instead: 143 (=11×13), 221 (=13×17). What connects both families is the same underlying reason: 2, 3, and 5 are small enough that people mentally check for them almost automatically, while 7, 11, and 13 sit in an awkward middle ground — large enough to slip past casual pattern recognition, yet still small enough to show up constantly in numbers people actually need to check by hand.
Why this matters beyond a party trick. Prime numbers are one of the few mathematical objects where being "almost right" gets you nothing — a number is either prime or it isn't, with no partial credit for being close. In contexts where primality actually matters (cryptographic key generation, for instance, where real prime numbers hundreds of digits long are tested using far more rigorous methods than mental shortcuts), a false positive on primality isn't a curiosity, it's a security failure. The 91 trap is a small, everyday reminder of a bigger truth: a quick heuristic is a starting filter, not a proof.
The reliable fix: always test all the way to the square root, not to wherever testing starts to feel sufficient. That cutoff isn't a convenience someone invented to save effort — it's the actual boundary past which no undiscovered factor can exist, so stopping short of it is a genuine gap, not a shortcut. For 91, stopping after 2, 3, and 5 (which might well feel like "enough" checking) skips directly past the one prime, 7, that actually breaks it.
A related myth worth debunking while we're here: some people assume 1 sneaks in as the "smallest prime." It doesn't, and the reason isn't a technicality — a prime needs two distinct divisors, and 1 only has the one (itself), so it's disqualified before the question of "how many other numbers divide into it" is even asked. Mathematicians don't just skip 1 out of convention; letting it count would quietly wreck the guarantee that every number above 1 has one specific, unrepeatable prime breakdown, since you could always toss in a few extra unnecessary 1's without changing the value.
Practicing the full method. The good news is that once you commit to always testing up to the square root — rather than stopping at whatever feels sufficient — the 91 trap and its relatives disappear entirely. It's not a harder method, just a more complete one: for any number under 100, you never need to test more than 2, 3, 5, and 7, since √100=10. This site's How to Check if a Number Is Prime guide walks through the complete method with several more worked examples, including a couple of the trickier trap numbers mentioned above. And if you just want a direct answer for a specific number without working it out by hand, every number from 1 to 550 has its own dedicated prime-check page here, showing the exact divisibility reasoning rather than a bare yes-or-no.
Building the habit that actually prevents the trap. The fix isn't memorizing a longer list of "trap numbers" — new ones keep existing beyond any list you could memorize. It's committing to the complete method every time: identify the square root of the number you're checking, list every prime up to that point, and test all of them, not just the ones that feel intuitively likely. For any number under 100, that's never more than four primes (2, 3, 5, 7); for any number under 200, it's never more than six (2, 3, 5, 7, 11, 13). That's a genuinely small, fixed amount of extra work in exchange for never falling into the 91 trap again.
A final worked example, to see the full method run cleanly on a similarly deceptive number. Is 187 prime? √187 ≈ 13.7, so test 2, 3, 5, 7, 11, 13. 187 is odd. Digit sum 1+8+7=16, not divisible by 3. Doesn't end in 0 or 5. 187÷7≈26.7, not whole. 187÷11=17 exactly. 187 = 11 × 17 — composite, caught by the second-to-last prime on the list, in a range plenty of people would have stopped testing before reaching.
The next time a number looks suspiciously prime — odd, doesn't end in 5, nothing obvious jumps out — it's worth remembering 91. Sometimes the number that "should" be prime by every quick glance is hiding exactly the kind of factor that quick glances are built to miss.