How the IQ Test Works
A plain account of what is in the test, how it is scored, and why the number you get is an estimate rather than a diagnosis.
The 50 questions
Every visitor gets the same 50 original questions in the same order. There is no question bank to draw from and no admin panel: the questions are fixed. They were written for this site, in the style of the reasoning puzzles that IQ tests are known for. None is copied from a commercial or clinical test.
| Area | Questions | What it asks of you |
|---|---|---|
| Numerical reasoning | 14 | Number sequences and quantity puzzles: spot the rule, then apply it. |
| Verbal reasoning | 13 | Analogies, word meaning and odd-one-out: how well you handle relationships between words. |
| Logical reasoning | 13 | Deduction, ordering and conditional reasoning: work out what must be true. |
| Visual patterns | 10 | Shape sequences and pattern grids: find the rule that completes the picture. |
Easy to hard
The exam runs from easy to hard: 20 easy questions, then 19 of medium difficulty, then 11 hard ones. Inside each level the four areas take turns, so no single skill dominates any stretch of the test. The early questions build confidence and the later ones separate stronger from weaker reasoning.
How your score is calculated
- Raw score. Each correct answer is worth one point, giving a raw score from 0 to 50. Blank answers score nothing.
- Estimated IQ. The raw score is placed on the standard IQ scale, where 100 is average and 15 points is one standard deviation: IQ = 100 + 15 × (raw score − 27) ÷ 8, rounded and kept between 55 and 145.
- Percentile. Using the bell curve, your estimated IQ is turned into the share of people it would sit above.
- Band. The estimate is matched to a descriptive band, such as “Average” or “Above average”. See the IQ score chart.
- Breakdown. Your correct answers are also totalled for each of the four areas, so you can see where you were strongest.
What different raw scores give
| Correct answers | Estimated IQ | Higher than about |
|---|---|---|
| 10 of 50 | 68 | 2% of test takers |
| 20 of 50 | 87 | 19% of test takers |
| 27 of 50 | 100 | 50% of test takers |
| 35 of 50 | 115 | 84% of test takers |
| 43 of 50 | 130 | 98% of test takers |
Why it is only an estimate
A real IQ score comes from comparing your answers with those of thousands of people, carefully sampled to represent a population. This test has no such comparison group. Instead it assumes that an average test taker answers about 27 of 50 correctly, with a typical spread of about 8 questions. Those two numbers are educated starting points, not measurements, and they can be recalibrated as real results come in.
Several other things also stand between a score and the truth:
- It is unsupervised. Noise, tiredness, distraction and help from a friend all change the result.
- The questions are the same every time, so retaking the test rewards practice.
- 50 questions is short. Full clinical tests use many more, across many more skills, and include timed tasks and memory.
- Scores near the top and bottom are the least reliable. A handful of questions decides the difference between very high and merely high.
Use the number as a talking point, not a label. To learn more about what IQ does and does not capture, read What is IQ?.
What happens to your answers
While you take the test, your answers are held in your browser so a refresh does not lose them. When you ask for your result they are sent to the server once, scored, and displayed. Nothing is stored, no cookie is set and no account exists. The privacy page has the details.