Skip to content
Purity Test Kit

The purity test for couples

The classic list asks what you have done. The couples version asks what the two of you have done — 50 statements that all start “have you two”, worth 2 points each, counting down from 100. Because you are both answering about the same shared history, the interesting output is not either score. It is the gap between them.

What a gap between your two scores is counting

Every answer moves a score by 2 points, so the difference between two results divides cleanly: a 6-point gap is three statements answered differently, a 10-point gap is five. Nothing else. The number cannot tell you which three, and it has no opinion about them.

When two people describe the same relationship and land on different totals, it is almost never because one of them lied. The usual reasons are duller and more useful to talk about:

  • Different thresholds. One of you counts a trip to a friend’s wedding as “a trip together”; the other was thinking of the week in another country.
  • Different memories. Items about arguments, secrets and money are the ones people most often remember at different sizes.
  • Different reading of “yet”. A statement half-true — you talked about moving in but have not — reads as yes to one person and no to the other.

That is the whole value of taking it separately: a 8-point gap points at four specific sentences you would otherwise never have compared.

What the 50 questions cover

Counted off the list itself. 30 of the 50 are about the relationship milestones themselves, which is what makes this version readable for a couple who have been together a month and a couple who have been together a decade.

Number of questions in each block of the 50-question couples test.
BlockQuestionsPoints at stake
Romance & dating3060
Sexual experience612
Online & modern life510
Alcohol & drugs36
Parties & nightlife24
Law & consequences24
Rule-breaking & mischief24

The 6 sexual questions sit in a single run at the end and are the reason the test carries an eighteen-and-over notice. They are about the two of you, not about anyone either of you knew before.

How to actually run it as a pair

  1. Answer on separate screens, in the same room. Reading each other’s ticks as they happen turns it into one shared answer sheet and destroys the only comparison worth having.
  2. Say the two numbers at the same time. Whoever speaks second otherwise adjusts, and everyone does it.
  3. Work backwards from the gap, not from the winner. Divide the difference by 2 and find that many sentences. There is no winner: a lower score is a longer shared history, which is what a couple is meant to accumulate.
  4. Stop when it stops being funny. The list includes items about cheating, lying about money and reading each other’s messages. Those are real conversations wearing a quiz format, and they do not have to be had tonight because a checkbox raised them.

Nothing is stored against either of you. A submission records which test it was, the score and the time — no account, no partner linking, no answers, unless you tick the optional box that shares which statements were checked as anonymous per-question counts. That box is off unless you turn it on, and it never attaches a set of answers to a person.

Comparing your pasts is a different test, and a riskier one

Couples usually arrive wanting one of two things. If you want to compare what the two of you have built, the couples version above is the one that measures it. If you want to compare who each of you was before, that is the classic 100-question test, taken separately — and it behaves differently in a way worth knowing before you start.

On the classic list the largest block is sexual history, and the score compresses all of it into one number that hides which items produced it. Two partners can be twenty points apart because of one long relationship on one side, or because of twenty separate items on the other. The score will not say, and the temptation to fill the silence with a guess is the part that goes wrong.

Neither test has recorded results here yet, so there is no median to quote. The statistics page fills in from the database as soon as there is something to read.

If it helps to know what any single number covers before you trade them, every score from 0 to 100 has its own page, and purity test vs innocence test explains why scores from two different question sets are not comparable at all.