What is here
A sixteen-item reasoning test across four domains — letter and number series, matrix reasoning, verbal reasoning and three-dimensional rotation — scored in your browser, with the error band printed next to the number. Around it sit three reference pages: how the test is built and scored, how to read a score, and who checked the explanatory content. That is the whole site. No course, no membership, no coaching, no app.
Why sixteen items and not sixty
Because the honest ceiling of an unsupervised browser test is low, and a longer test mostly disguises that. Sixteen items with a published error band tells you roughly where you sit and admits how roughly. Sixty items with no error band tells you a number and implies a precision nobody has established.
The test used to be longer. It carried thirty-three items, then twenty-one plus two timed tasks and a five-index profile. Each version reported more and knew less. The current one reports a total, four domain counts and a ±13 interval, and every one of those figures has arithmetic behind it that we publish.
Five things we will not publish
Deciding what to leave out shaped this site more than deciding what to include. Each of these is a reliable traffic source in this niche and each is off the table.
- Celebrity IQ lists. Almost every figure in circulation is invented, back-calculated from childhood anecdotes, or attributed to people who were never tested. Publishing them is fabrication with a table around it.
- National IQ rankings. The datasets behind these have known problems with sample sizes, sampling methods and estimated entries. They travel well and inform nobody.
- Claims that a product raises IQ by a stated number of points. Whether training on one cognitive task transfers to untrained ones is genuinely unsettled, and a specific promised gain goes well past the evidence.
- Anything framed as diagnosis. No page here will tell a reader they are gifted, have a learning difficulty, or should be worried about their cognition. That belongs with a clinician.
- A paid tier. No locked report, no certificate, no “professional” version. Once there is something to sell, every sentence about accuracy has a second motive.
Free, stated once
Nothing is asked of you at any point: no email, no account, no card. There is no mechanism by which a charge could appear, because there is no payment path on this site at all — which is worth stating against the trial-to-subscription pattern documented in public reviews of several competing services.
Scoring itself runs in your browser. Your individual answers are never transmitted to us and could not be produced if you asked for them. Everything that does move is itemised in the record of what we hold, and the cookie page exists mainly to say that we set none at all.
What we publish alongside the score
Three things that most free tests leave out, and the reasons they get left out are instructive.
- The margin of error. A number without one invites false precision, and false precision is what makes a score feel like a verdict.
- The point at which the result stops meaning anything. A raw score of five or under is inside what guessing produces on a six-option test, and the page says so rather than printing a figure.
- The worked reasoning for every item. Sixteen questions cannot measure anyone accurately. Sixteen explanations are worth having regardless, and they are the part of the output we would defend hardest.
Where the money comes from
Costs are covered by a small number of contextual links and partnerships, marked where they appear. No partner sees content before it publishes or has any say over the test, the scoring, or what we write about accuracy. The arrangement only works while those two things stay separate; if they ever conflicted, the partnership is the part that goes.
Who checks what
Pages are written in-house. The explanatory content was reviewed by Dr Mark Ashton Smith, a cognitive neuroscientist at the University of Essex Online, on 22 July 2026 — a review that covered score interpretation and the treatment of measurement error, and did not extend to authoring items or validating the instrument. His page discloses that he runs a commercial cognitive training business.
The test was rebuilt after that review, so the current content has not been re-reviewed. The review log on his page records exactly which version he saw. We would rather carry a stale review date honestly than move it because the text was edited.
The legal pages, the contact page and this one are ours alone and carry no review credit, because claiming one would be padding.
When we get it wrong
Fees change, editions get replaced, thresholds move, and sometimes an item turns out to have two defensible answers. Our rules for that are short:
- Checkable claims cite a primary source — the paper, the publisher's own documentation, the organisation's stated policy — never another site repeating it.
- Figures carry the date they were last verified.
- Complaints from review platforms are labelled as user reports, not presented as established fact.
- A review date moves only when a review actually happens, never because the text was edited.
- Substantive corrections are made and noted rather than quietly overwritten.
If something here is wrong, say so. Being corrected costs an afternoon; being quietly wrong costs the point of the site.
What this is not
IQTest is an educational publisher, not a psychology practice and not a clinical service. It is not affiliated with Mensa, Pearson, the University of Essex or any test publisher named on this site. Nothing here diagnoses anyone or qualifies anyone for anything, and a supervised assessment is the only route to a score that carries weight anywhere. The site terms put that in concrete form, with a list of the specific purposes a result from here cannot serve.