Free Top 2% Challenge

What a top-2% test actually asks of you

A test built for the very top of the distribution does not ask what you have learned. It asks how quickly you can build a rule out of material you have never seen and then apply that rule under time pressure. This challenge is written in that spirit. Every one of the thirty items here was authored for this page, and every one is deliberately harder than the questions on a general-population IQ test, because the interesting part of the distribution — the top two per cent — is compressed into a very narrow band of raw scores. A test made of easy questions cannot separate people inside that band at all.

The four question families

The test rotates through four families so that no single strength can carry you. Figure series show four abstract frames and ask for the fifth; the governing rule may be a rotation step, a growing count, or a progression in the number of sides of a polygon, and the distractors are each a single controlled deviation from the correct figure. Number sequences are built on layered rules rather than plain arithmetic — alternating offsets around squares, cubes minus a constant, second differences that are themselves square numbers, or a doubling map with a fixed correction term. Word logic covers analogies of degree, hidden anagrams, letter ciphers, syllogisms and small ordering puzzles. Spatial folding asks which faces of an unfolded cube net end up opposite one another, and how many holes appear when a folded sheet of paper is punched and opened out.

Why the timer matters

Thirty minutes for thirty questions is not generous, and it is not meant to be. At this level of difficulty, almost everyone who can solve a given item can solve it eventually; the discriminating variable is how long it takes. Supervised admission tests exploit exactly this, which is why they are timed tightly rather than made impossibly obscure. If you routinely finish this test with ten minutes to spare and still score in the top band, you are working at the speed that a real sitting expects. If you run out of clock with six questions untouched, speed rather than reasoning is what is costing you, and that is the more trainable of the two.

Reading your score

Your raw total is mapped onto the familiar scale where 100 is average and fifteen points make one standard deviation. Because every item here is hard, the mapping is compressed at the bottom and generous at the top compared with an ordinary test: a middling raw score here still corresponds to a perfectly respectable ability level. Treat the number as a band of roughly ten points rather than a precise value.

How to prepare properly

The single largest gain from practice is the removal of format shock. Candidates who have never seen a cube net or a second-difference sequence lose two or three minutes simply working out what is being asked, and those minutes are the difference between a qualifying and a near-miss score. Beyond that, three habits help. First, name the rule out loud before you look at the options — distractors are designed to look correct to someone who has not committed to an explanation. Second, when a number sequence resists you, write the differences, then the differences of the differences; two levels are almost always enough. Third, budget strictly: if an item has not yielded in ninety seconds, guess, move on, and come back with the Back button if the clock allows.

How this differs from our other tests

The culture fair IQ test uses only shapes, so it is the better instrument if English is not your first language or if you want an estimate uncontaminated by vocabulary. This test deliberately mixes verbal and spatial material because real admission tests do, and because a mixed profile tells you more: a large gap between your verbal and figural performance is itself a finding worth knowing. Remember that no online test, this one included, counts as formal evidence anywhere. It is a rehearsal, and its job is to tell you whether booking a supervised sitting is worth your time.