Atlas/IQ/Full-Scale IQ Test: 72 Items and a Routed Extreme Block
Long-form reasoning tests
Full-scale IQ test: 72 items, a routed extreme block, and the arithmetic
IQ test with a full scale to 160 — estimated range, formula shown. 72 timed items in five tiers, 20 extreme items if you earn them, every rule published.
Free, no account
92 items · 55 min
Answers stay in this tab
Every answer shown afterwards
What this assessment measures
IQ test with a full scale to 160 — estimated range, formula shown. Seventy-two original items run in five difficulty tiers on one 55-minute clock: figure matrices carrying two and three simultaneous rules, set-logic grids, abstract analogies, quantifier-chain syllogisms, interleaved number series and composed rotations. Solve at least 56 of the 72 and the test adds a block of twenty extreme items, because below that point the core alone cannot tell 145 from 160. You get a MyTestAtlas score out of 100 and a 68 percent range on the IQ scale, produced by a three-parameter item model whose every constant is on the method page. It is an estimate anchored on difficulties we assigned ourselves, not a measured IQ.
What this test looks at
figure matrix · 33 items
spatial · 17 items
verbal reasoning · 23 items
numerical reasoning · 19 items
Time
55 minutes
Items
72 + 20 routed
Result
attainable-range pattern summary
Source
Original MyTestAtlas items and deterministic figure generator
Material limit
Ninety-two browser items with no norm group, no supervision, no age or language adjustment and no control over your conditions produce an estimate, not a measured IQ. The band's width is computed; its position rests on two constants we chose — and at the top of the scale a plausible error in them is worth more than the band. The range is shown on the first completed attempt within 55 minutes only, stops at 160, and must never be represented as an IQ score to an employer, a school, a society or a clinician.
This is the full-scale form: 72 items on one 55-minute clock, plus 20 extreme items offered at 56 of 72 correct. The 36-item quick form takes about 20 minutes and is clamped to 80–135.
This route calculates its result in this browser tab; answers and result text are not sent as analytics properties or written to browser storage.
Your progress
Timing has not started.
Review
You answered every item
Your answers are still only in this tab. Calculate the result now, or go back and change them.
Your result
—
Your map
How this range is made → Estimate from a published formula with author-set anchors, not a measured IQ. Shown on the first attempt within 55 minutes.
Review every item, your answer, and its rule family
Your answers were not sent or stored. Restart, or contact support with error code MTA-LOCAL-1—not your answers.
Use the result
Why this route is different
Free tests that advertise a high ceiling almost never say how many items sit up there, and the arithmetic says you need between 25 and 44 of them close to the target to place anyone near 160 with a margin of five points. This test has twenty, it routes them only to the people they measure, and it prints the margin it actually achieves rather than a number with the uncertainty removed.
A small next step
Open the answer review, find the tier where your accuracy fell off, and redo three of those items on paper — writing the rule in one sentence before choosing an option, which is the habit the composed items are built to reward.
Questions people ask
Can this test give a 160?
Yes, and here is what that does and does not mean. An IQ of 160 is four standard deviations above the mean, about one person in 31,600. Placing someone there within about five points needs roughly 25 to 44 items pitched at that level, which is what the twenty-item extreme block is for, and we print the margin we actually achieve. What we cannot do is the other half of the job: a real 160 comes from comparing you with a representative sample, and a sample large enough to contain people at 160 does not exist on this site and will not — it would take something like 114,000 finished tests to observe thirty of them. So our 160 is what our formula returns for a pattern of answers, calibrated against difficulties we set. Treat it as a strong result on a hard test with the method published, not as a measured IQ.
How is this different from the 36-item quick form?
The quick form takes twenty minutes and is clamped to 80 to 135, because thirty-six four-option items carry too little information to resolve anything outside that. This one runs 72 items in five tiers with option counts rising to eight, which cuts the guessing floor from one in four to one in eight where it matters, and it can resolve results up to 160 once the routed block is included. Both publish the same formula; this one has the items to back a wider scale.
What is the extra block and how do I get it?
If you solve at least 56 of the 72 core items, the test adds twenty items pitched far above the rest — items a person of average ability solves at about the guessing rate. It is offered rather than imposed, and you can finish without it. Without those items the estimate stops at 145 and says so, because the core alone cannot separate a result above 145 from one at it.
Why a 55-minute clock with no extra time?
A score only means something when everyone is given the same conditions, and the clock is the one condition we can hold. Fifty-five minutes for 72 items is about three quarters of a minute an item, a little more generous than the 35 items in 25 minutes that Mensa Norway publishes for its practice test. Run past the cap and you still get the MyTestAtlas score and the full answer review, but no range, and the result says why.
Why a range and never a single number?
Because a single number would hide how much of it is measurement. We print a 68 percent band whose width comes from the information the items you saw actually carry at your level, and we print the wider 95 percent band beside it. We also publish what happens to the same result if the two constants we chose are wrong, which at the top of the scale is worth more than the band itself.
Is it free, and does it need an account?
It is free, there is no account, no email and no paywall. Your answers stay in this browser tab and are never written to storage or sent anywhere. Closing the tab loses the result, which is the price of not collecting anything.
Source and method boundary
These items and their scoring are original to MyTestAtlas. The route is labelled as practice, reflection, or entertainment and has no clinical, IQ, hiring, or population-norm interpretation.