The My Path IQ Test is a free 40-item online cognitive assessment covering four reasoning domains — verbal, numerical, pattern recognition, and spatial reasoning. The test is not timed (we score for accuracy, not speed); takes 25–35 minutes. Results show your scaled IQ score against the standard normal distribution (mean 100, SD 15), a percentile against a representative reference sample, and a section-level breakdown so you can see whether your reasoning strength is even or skewed toward specific domains.
Free to take. No credit card required.
Our IQ instrument is calibrated against the standard normal IQ distribution (mean 100, SD 15) using item-response-theory (IRT) scoring on a representative reference sample. The test covers the four reasoning domains that load most heavily on the general intelligence factor (g) in published cognitive batteries — verbal reasoning, numerical reasoning, pattern recognition, and spatial reasoning.
Each item is calibrated for difficulty using item-response-theory (IRT) — items aren't worth equal points; harder items contribute more to the score when answered correctly. Your raw response pattern is run through the IRT model to produce a latent-ability estimate, which is then transformed onto the standard IQ scale (mean 100, SD 15). Scaled IQ scores in the 85-115 range cover ~68% of the population; 70-130 covers ~95%.
Verbal — analogies, vocabulary in context, and verbal logic puzzles. Numerical — number sequences, mental arithmetic, and quantitative reasoning. Pattern recognition — visual abstraction tasks similar to Raven's Progressive Matrices. Spatial — mental rotation, spatial transformation, and 3D-from-2D inference. The four domains are highly intercorrelated (r ≈ 0.5-0.7) — that intercorrelation IS the general intelligence factor (g) — but the section-level breakdown surfaces real differences between users.
Reference: Carroll, J. B. (1993). Human Cognitive Abilities: A Survey of Factor-Analytic Studies. Cambridge University Press.
Most IQ tests time-pressure the user, which means the score blends raw reasoning ability with processing speed. We deliberately don't time — our scoring rewards accuracy over speed. This produces a slightly higher mean score than timed tests for the same population (about 5-7 points higher in our reference data), so we calibrate the scale to that elevated mean. Your untimed-IQ percentile is meaningful WITHIN our scoring framework but isn't directly interchangeable with timed-IQ percentiles from clinical instruments like WAIS-IV.
Cognitive ability is one of the most-replicated predictors of academic outcomes, occupational complexity capacity, and certain life outcomes — the meta-analytic evidence base is enormous. But it does NOT predict: emotional regulation, work ethic, social functioning, creativity in unstructured domains, or moral reasoning. People who treat their IQ score as a global verdict on their worth are reading the construct wrong. The score answers a narrow question: how well do you reason on this specific battery of structured items.
Reference: Schmidt, F. L., & Hunter, J. E. (1998). The validity and utility of selection methods. Psychological Bulletin, 124(2).
No — and we're explicit about that. Clinical IQ instruments (WAIS-IV, Stanford-Binet) are administered 1-on-1 by a trained psychologist, take 60-90 minutes, and produce scores meaningful for clinical or educational decisions. Our online test uses item-response-theory and a representative reference sample, but lacks the clinical observer's qualitative input. As a free benchmarking tool: solid. As an evaluation for clinical or legal purposes: not a substitute.
Most likely the test is in its expected ±5-7 point margin of error, and your expectation was off. Self-estimated IQ correlates only modestly with measured IQ (~0.3) — most people's self-estimate is anchored on social comparisons that are noisy. If your score deviates from a previous IQ test result by more than 10-15 points, the most common cause is test-context difference (timing, alertness, motivation) rather than real ability change.
Yes, but in a specific, somewhat misleading way. Repeated administration of the same items or item types produces score gains (called 'practice effects') of 5-10 points, which are NOT genuine cognitive ability gains — they're test-format familiarity. Genuine cognitive ability is much harder to move; deliberate practice of working memory and processing speed produces small effects (2-3 points) over months of training. Don't take this test more than once every 30 days if you want a stable longitudinal measure.
Most people have noticeable section-level skew — high verbal + low spatial, or strong numerical + weak pattern recognition. The skew is informative: it predicts subject-area fit (e.g. high spatial → engineering, architecture; high verbal → law, writing) more than the single overall score does. The 4-section breakdown is one of the most useful outputs of the test.
IQ is one of the strongest single predictors of job performance across all occupations — meta-analyses put the correlation around 0.5 for cognitively-complex jobs and 0.3 for less-complex jobs. But it's not the only predictor: Conscientiousness (Big Five) is comparably strong, and integrity / motivation / EQ add incremental validity. Your IQ score sets a kind of capacity ceiling for cognitive complexity; whether you reach that ceiling depends on the other factors.
Different cognitive abilities follow different lifespan trajectories. Crystallized intelligence (verbal knowledge, accumulated reasoning patterns) keeps growing into your 60s. Fluid intelligence (raw pattern recognition, novel problem solving) peaks in your 20s and slowly declines from there. The single 'IQ score' averages these trajectories — overall IQ is fairly stable from age 20-60 in healthy adults, with minor declines after 60 driven mostly by the fluid component.
Most likely no. Average IQ (85-115) covers ~68% of the population and is fully sufficient for the vast majority of life and work demands. Below 85 is statistically below average but is not clinically meaningful unless paired with adaptive functioning challenges. The IQ score answers one specific question (cognitive reasoning on these items); it does not answer "am I capable?" or "am I worthy?". Take the section-level breakdown as more useful than the single number for planning around your strengths.
All cognitive tests have some cultural and linguistic loading — the verbal section relies on English-language fluency, and pattern items can favor people who've seen similar structures before. Our tests are translated and reviewed for cultural fit in 6 locales. We're explicit that an IQ score is meaningful relative to the test's reference sample, not as an absolute measure of innate ability. For users in non-Western or non-formal-education contexts, the score is best interpreted as "how does my reasoning fit the patterns this test rewards" rather than "how smart am I in general".
Your IQ report. Scaled score, percentile, section-level breakdown, and a calibrated narrative — without the bombast of "genius" / "prodigy" framings that other free tests use.
Your single-number IQ score on the standard scale (mean 100, SD 15), with the corresponding percentile against a representative reference sample. Includes the ±5-7 point margin of error explicitly.
Verbal / numerical / pattern recognition / spatial reasoning percentile bars. The skew across sections is often more informative than the overall score for understanding your reasoning profile.
For paid users: see exactly which items you got right vs. wrong, with the correct answer and a brief explanation of the reasoning step you missed. Useful for test-prep contexts and for users who want to understand their error patterns.
A 2-3 paragraph interpretation of your profile that names the section skew explicitly, references the score band's practical implications, and avoids the "genius" framing common in free-tier IQ tests.