Free Fluid Intelligence Test

What a fluid intelligence test actually measures

Psychologists split reasoning ability in two. Crystallised intelligence is everything you have already banked: words you know, facts you can recall, procedures you were drilled in at school. Fluid intelligence is what is left when you take all of that away — your capacity to meet a problem you have never seen, find the rule hiding inside it, and apply that rule correctly under a clock. This test is built to isolate the second thing. There is not a single word to define, not a single sum to compute, and nothing that rewards having read more books than the person sitting next to you.

How the 22 items are constructed

Every item is written by hand from an explicit rule rather than borrowed from an existing question bank. Six of them are figure analogies: the first figure changes into the second by one or two controlled steps, and you apply the same change to a third figure. Six are rule series, where a row of four figures moves through a progression of count, shading, rotation or size and you supply the fifth. The remaining ten are progressive matrices — a three by three grid where one attribute varies down the rows and another varies across the columns, and the bottom right cell is missing. In every case the five wrong options are single deliberate changes to the correct one, so eliminating options by feel is genuinely hard, and in every case exactly one option satisfies the rule.

Why the questions get harder as you go

The early items vary one attribute at a time: shading flips, a count grows, a figure shrinks. From roughly the middle onward two attributes move at once, and the last few require you to hold a rule for rows and a different rule for columns in mind simultaneously. That is not decoration — the number of independent rules a person can track at once is the single feature that separates easy matrix items from hard ones, and it is what the score is really counting. If you feel the difficulty step change around item twelve, the test is working as intended.

How your score is calculated and read

Your raw total out of 22 is placed on the standard scale, where 100 is the population average and each 15 points is one standard deviation. Because 22 items is a short test, treat your number as the centre of a band roughly ten points wide rather than as an exact value. The table below is the ordinary interpretation of each range.

How to get a number you can trust

• Take it once, rested, and not as your fourth test of the evening. Fluid reasoning is the first thing fatigue takes away and the last thing it gives back.
• Answer everything. Nothing is deducted for a wrong answer, so a considered guess always beats a blank cell.
• If an item resists you for more than a minute, mark your best guess and move on. You can return with the Back button while the clock still allows it.
• Use the largest screen you have. Squinting at a matrix on a phone costs seconds that the timer does not give back.
• Do not look for the answer anywhere. The moment you search, you are measuring your search skills instead of your reasoning.

How this differs from the other tests on the site

The culture fair IQ test is the closest relative: it also uses only shapes, but every item there is a three by three matrix generated from a rule family, so it measures one format very steadily. This test deliberately spreads the same underlying ability across three formats, which makes it harder to game by getting good at matrices specifically. The quick IQ test mixes in verbal analogies and number series, so it partly measures crystallised knowledge and is not comparable. If your fluid score sits well above your quick score, the usual explanation is that your reasoning outruns your vocabulary or your arithmetic fluency; if it sits well below, you are probably being slowed by the clock rather than by the problems.