🧩

Non-Verbal Reasoning Test

Spot the pattern — no words, no arithmetic.

Question

—

Correct

Best

—

🧩

15 pattern puzzles, no words

Matrices, sequences and odd-one-out — the three shapes that make up almost every non-verbal reasoning paper. Every puzzle is generated fresh, so you can practise as often as you like.

About the Non-Verbal Reasoning Test

Non-verbal reasoning measures how well you spot a rule in a set of shapes and apply it — without language or arithmetic getting in the way. It is the format behind 11+ non-verbal reasoning papers and the abstract or inductive reasoning sections of graduate aptitude tests. Every puzzle here is generated fresh rather than drawn from a fixed bank, so you can practise as many papers as you like and never repeat one.

How to play

Pick a difficulty and press Start. You get 15 puzzles of three kinds.
A matrix asks which figure completes a 3×3 grid; a sequence asks what comes next; odd-one-out asks which figure breaks the rule.
Check one property at a time — shape, number, angle, shading — rather than judging the figure as a whole.
After each answer the rule is spelled out, so a wrong answer still teaches you the pattern.

What it measures

This test focuses on your abstract pattern recognition (fluid reasoning). Scoring 11 or 12 out of 15 on medium means you are reliably reading two rules at once. Real papers are usually tighter on time than they are on difficulty — accuracy first, then speed.

Why play it

Practises the exact format used in 11+ and graduate abstract reasoning papers
Puzzles are generated, so you can sit unlimited papers without repeats
Every question explains its rule afterwards, which is where the improvement comes from

Did you know?

Non-verbal reasoning is used in selection precisely because it leans less on vocabulary and schooling than verbal tests do.
The 3×3 matrix format was popularised by Raven’s Progressive Matrices, published in 1938 and still one of the most widely used measures of abstract reasoning.
Most matrix puzzles combine two independent rules — one across the rows, one down the columns.
The commonest mistake is solving one property and answering immediately, which is exactly what the wrong options here are built to catch.

What counts as a good score?

No published norms exist for this format, so the ranges below are practical skill tiers rather than measured population data.

Level	Score out of 15 (medium)	What it means
Getting started	0–7	Usually a method problem rather than an ability one: judging the figure as a whole instead of checking each property in turn.
Developing	8–10	You are finding single rules reliably. The gap to the next band is almost always the second rule.
Strong	11–13	Reading two independent rules at once — a row rule and a column rule — which is what most matrix items require.
Excellent	14–15	Comfortably above what these papers are pitched at. Move to hard, where the rules step in twos and the shape pool widens.

These are practical tiers for this generator, not norms for any published test. Real 11+ and graduate papers are standardised against a specific cohort and scored against that cohort — a score here tells you how you are progressing, not what you would get on the day.

How to improve your non-verbal reasoning score

1.Check one property at a time

Shape, then number, then angle, then shading. Almost every wrong answer comes from spotting one rule and answering before checking whether a second is running alongside it.

2.Say the rule out loud

"Across, the count goes up by one; down, the shape changes." If you cannot put the rule into a sentence, you have not found it yet — you are pattern-matching on appearance.

3.Predict before you look at the options

Work out what the missing figure must be, then find it. Reading the options first invites you to rationalise a plausible-looking wrong one, which is exactly what they are built to do.

4.Use the options as a check, not a crutch

If two options differ only in shading, shading is one of the rules and worth checking first. The answer set tells you which properties matter.

5.Accuracy before speed

Real papers are tight on time, but speed built on a shaky method collapses under pressure. Get the method automatic first; the pace follows.

Where this format comes from

The 3×3 matrix was popularised by Raven's Progressive Matrices, published by John C. Raven in 1938. The design goal was a measure of reasoning that leaned as little as possible on language, schooling or cultural knowledge — which is why it spread so widely, and why versions of it still appear in both school entrance exams and graduate selection nearly a century later.

The same format carries different names depending on who is administering it. Schools call it non-verbal reasoning; employers and test publishers call it abstract, inductive or diagrammatic reasoning. The underlying task — infer a rule from figures, apply it — is the same one.

What it does and does not tell you

Matrix-style tests correlate well with general reasoning ability, which is why they are used in selection. But they are also demonstrably trainable: scores rise substantially with practice, and much of that gain is method rather than raw ability — knowing to check each property in turn is worth several marks on its own.

That cuts both ways. It means practice genuinely helps, which is the honest case for a page like this. It also means a practice score is not an ability measurement, and anyone telling you a free online test has diagnosed your child's reasoning ability is overselling it. Use it to build method and confidence with the format.

Why these puzzles are generated, not copied

Every item here is constructed from rules at the moment you press start, rather than drawn from a fixed bank. That means unlimited practice with no repeats — the usual limit on free resources in this niche, which mostly offer a handful of downloadable papers.

It also imposes a constraint worth knowing about: a generated puzzle is only useful if it has exactly one defensible answer. The generator enforces that structurally — the row and column rules never touch the same property, wrong options differ from the answer in exactly one property, and rules that would be invisible on screen (rotating a circle, or a square through 90°, which looks identical) are excluded rather than shipped.

More memory tests

🔢 Number Memory Test 🧩 Visual Memory Test 🐵 Chimp Test 💬 Verbal Memory Test 🔢 Schulte Table 🎨 Stroop Test ⚡ Reaction Time Test 🔁 Backwards Digit Span Test 🎯 N-Back Test (2-Back) 🎯 Aim Trainer ⌨️ Typing Test 🔢 Sudoku 🔢 2048 💣 Minesweeper 🧩 Nonogram ➕ Kakuro 🔤 Word Guess 🧩 Sliding Puzzle 🃏 Solitaire

Prefer a themed game?

🃏 Card Matching Games 🔁 Sequence Memory Game 🔍 What's Missing? Game 🔎 Odd One Out Game 🧐 Spot the Difference Game

Frequently Asked

What is a non-verbal reasoning test?

A test of your ability to spot and apply rules in patterns of shapes, with no reading or arithmetic involved. It typically mixes matrices, sequences and odd-one-out, and appears in 11+ entrance exams and in graduate aptitude batteries under names like abstract, inductive or diagrammatic reasoning.

Are these real 11+ or aptitude test questions?

No — and deliberately so. Every puzzle is generated by rule rather than copied from any paper, which means unlimited fresh practice with no repeats. The formats match what the real papers use; the individual items are our own.

How do I get better at non-verbal reasoning?

Check properties one at a time — shape, number, angle, shading — instead of judging figures as a whole. Almost every wrong answer comes from spotting one rule and stopping before checking the second.

Is this the same as an IQ test?

It uses the same format as the matrix sections of some IQ batteries, but it is not an IQ test. A real IQ score is standardised against a matched cohort under controlled conditions. Treat this as practice with the format, not a measurement.

Can you actually improve at non-verbal reasoning?

Yes, measurably — scores rise with practice, and a good part of that is method rather than raw ability. Learning to check each property in turn instead of judging the figure as a whole is worth several marks by itself.

Which difficulty should I use for 11+ practice?

Medium is the closest fit for most 11+ non-verbal reasoning papers. Easy is a good starting point for building the checking habit; hard steps rules in twos and widens the shape pool, which is closer to graduate-level abstract reasoning.

Why do some questions look almost identical?

The wrong options are built by changing exactly one property of the correct answer. That is deliberate: it means a half-solved pattern lands on a plausible wrong answer, which is precisely how real papers separate careful reading from guessing.

References

Raven, J. C. (1938). Progressive Matrices: A Perceptual Test of Intelligence.
Raven, J. (2000). The Raven's Progressive Matrices: Change and stability over culture and time. Cognitive Psychology, 41(1), 1–48.
Hayes, T. R., Petrov, A. A., & Sederberg, P. B. (2015). Do we really become smarter when our fluid-intelligence test scores improve? Intelligence, 48, 1–14.