IQ Test for Kids: Free, Age-Normed, Instant Results
Our 35-puzzle matrix test uses no words or numbers, so children who can't yet read fluently can still take it. Pick your child's age group — scoring norms adjust automatically — and the result appears instantly, free, with no email.
Age-specific norms, what to expect, and how scores at this age should be read →
Age-specific norms, what to expect, and how scores at this age should be read →
Age-specific norms, what to expect, and how scores at this age should be read →
Age-specific norms, what to expect, and how scores at this age should be read →
Age-specific norms, what to expect, and how scores at this age should be read →
Go straight to the test and choose an age group on the start screen →
How scoring adapts to age
Children aren't small adults on a matrix test: the same raw performance means very different things at 7 and at 15. Our test handles this the way professional instruments do — the puzzles stay identical, but the score is computed against age-group norms. Here is the raw performance (out of 35) that maps to key IQ levels under our provisional norms:
| Age group | IQ 85 | IQ 100 (average) | IQ 115 | IQ 130 (gifted) |
|---|---|---|---|---|
| Ages 6–7 | 3 | 8 | 13 | 18 |
| Ages 8–9 | 6 | 11 | 17 | 22 |
| Ages 10–11 | 8 | 14 | 20 | 26 |
| Ages 12–13 | 10 | 16 | 22 | 28 |
| Ages 14–16 | 11 | 17 | 23 | 29 |
Provisional calibration — see methodology. Professional tests (WISC-V, WPPSI-IV, Stanford–Binet 5) use norms standardized on thousands of children per age band; an online screening test cannot match that precision.
What professional testing looks like (and when it's worth it)
Schools and psychologists use individually administered batteries — most commonly the WISC-V for ages 6–16 and the WPPSI-IV below that. A qualified examiner works one-on-one with the child for 45–90 minutes and produces a full cognitive profile: verbal comprehension, visual-spatial reasoning, fluid reasoning, working memory and processing speed. Gifted programs typically look for a full-scale score of 130 or above, though many districts also accept high subscale scores or use screening tests like the CogAT first.
An online test is the wrong tool for those decisions — but it is a reasonable first signal. If your child scores far above their age norms here, professionally administered testing is worth the cost when a concrete decision (program entry, acceleration, IEP support) hangs on the result.
Frequently asked questions
How accurate is an online IQ test for a child?
Less accurate than for adults, and parents should know that upfront. Children's scores vary more day to day, attention wanders, and our age norms are provisional calibrations rather than clinical standardizations. Treat the result as a rough screening estimate — useful for curiosity and as a first signal, never for decisions. School placement and diagnosis require a professional assessment (typically the WISC-V).
What is a good IQ score for a child?
The scale is the same as for adults: 100 is average for the child's age group, 90–109 is the average band, and 130+ (roughly the top 2%) is the conventional gifted threshold used by most school programs. What differs by age is the raw performance expected to reach those scores — which is what our age norms adjust for.
When should I get my child professionally tested?
Three common triggers: a school gifted program requires a qualifying score; there are learning difficulties where a cognitive profile would guide support (an IEP or 504 evaluation); or a big mismatch between evident ability and school performance needs explaining. A psychologist-administered WISC-V costs a few hundred dollars, takes about an hour, and produces subscores an online test cannot.
Can my child practice to improve their score?
Familiarity with matrix puzzles reliably adds a few points — which is precisely why schools discourage heavy test prep before gifted screening: it inflates the entry score without changing the underlying ability the program is trying to serve. Light exposure so the format isn't alien is reasonable; drilling is self-defeating.