UCAT Scores Explained: Deciles, Bands and the IQ Question

A good UCAT score depends on the university. The UCAT Consortium says no single score guarantees selection and not every university sets a minimum. For scale, the 2025 mean total was 1891 out of a possible 2,700 across 41,354 tests, the 5th decile was 1880, and a total of 2100 was the 80th percentile. The 2026 figures are still preliminary. No source we found calls the UCAT an IQ test or converts a UCAT score into an IQ, and this guide explains why such a conversion would mislead.

How is the UCAT scored in the 2026 cycle?

The 2026 test has four subtests, according to the Consortium's test-format page. Verbal Reasoning has 44 questions in 22 minutes, Decision Making 35 in 37 minutes, Quantitative Reasoning 36 in 26 minutes and Situational Judgement 69 in 26 minutes. Each of the three cognitive subtests is scored from 300 to 900, giving a total of 900 to 2700, and Situational Judgement is reported as a band from 1, the highest, to 4. There is no negative marking. Abstract Reasoning was removed in 2025, and the Consortium says scores from 2025 onward are not directly comparable with earlier years, when the total ran from 1,200 to 3,600 across four cognitive subtests.

Not every question counts. The 2025 technical report, which the Consortium links publicly, shows that each subtest carries unscored trial items: Verbal Reasoning has 10 scored testlets of four items plus one unscored testlet, Decision Making has 31 scored items plus four unscored, and Quantitative Reasoning has eight scored testlets of four items plus one unscored. That is 40, 31 and 32 scored items out of the 44, 35 and 36 a candidate sees. The Situational Judgement bands are set by score ranges: in 2025 Band 1 covered scaled scores of 658 to 900, Band 2 596 to 657, Band 3 502 to 595 and Band 4 300 to 501, with intended shares of 22, 38, 30 and 10 per cent.

What do the latest UCAT statistics show?

The Consortium publishes a mean and a decile table for each test year, and defines a decile simply: the 1st decile is a score at the 10th percentile, the 2nd decile a score at the 20th percentile, and so on. Here are the total-score deciles for 2025, final, based on 41,354 tests up to 26 September 2025.

• 1st decile: 1580
• 2nd: 1680
• 3rd: 1760
• 4th: 1820
• 5th: 1880
• 6th: 1950
• 7th: 2010
• 8th: 2100
• 9th: 2220

And the preliminary deciles for 2026, based on 20,794 tests up to 13 September 2026, with the Consortium expecting about 22,000 more candidates.

• 1st decile: 1630
• 2nd: 1750
• 3rd: 1840
• 4th: 1910
• 5th: 1980
• 6th: 2040
• 7th: 2110
• 8th: 2190
• 9th: 2300

The 2025 final mean total was 1891, with a standard deviation of 244.84 in the technical report, and subtest means of 602 for Verbal Reasoning, 628 for Decision Making and 661 for Quantitative Reasoning. The 2026 preliminary means are 1970 for the total and 614, 657 and 699 for the subtests. Do not read that as a rise. The Consortium says the 2026 scores should be treated with significant caution and gives its own example: the 2025 preliminary mean was 1939 and the final mean 1891. Situational Judgement shifted too. In 2025, 21 per cent of candidates were in Band 1, 39 per cent in Band 2, 29 per cent in Band 3 and 10 per cent in Band 4; the 2026 preliminary shares are 16, 42, 32 and 10. The technical report says Verbal and Quantitative Reasoning scores were downscaled by 10 points in 2025 and the band cut-offs revised, so bands are not like-for-like with 2024, when Band 1 held 13 per cent.

What is a good UCAT score?

The Consortium says universities use UCAT results in different ways: some set a threshold score, some weight it alongside academic results, and some use a cut-off for a subtest or for the Situational Judgement band. Its FAQ says there is no single score that guarantees selection and not all universities have a minimum. One published rule: the University of Edinburgh says its minimum UCAT total cut-off for 2027 entry is 1850, that it then ranks scores and divides the cohort into deciles, and that applicants in Situational Judgement Band 4 will not be considered. It adds a 5 per cent uplift to the UCAT total for Flag applicants and 10 per cent for Plus Flag and Scottish SIMD40 applicants. Newcastle says it cannot predict its threshold because it depends on the number of applicants and their average score.

Set Edinburgh's 1850 against the national table and it is not a high bar. On a normal curve with the 2025 final mean of 1891 and standard deviation of 244.84, 1850 is about the 43rd percentile, between the 4th and 5th deciles; against the 2026 preliminary table it falls between the 3rd and 4th. A minimum is a floor for consideration, and the ranking that follows is what counts. That is our comparison, and a floor at one university says nothing about the others.

Is the UCAT an IQ test?

No source we found says so. The Consortium describes the test as assessing a range of mental abilities identified by medical and dental schools as important, and its technical report says it is not an exam of student achievement and holds no curriculum or science content. The authors of a large validity study, McManus and colleagues (BMC Medicine, 2013), write that the UKCAT aims to assess fluid intelligence and other attributes thought to be important in decision making, while specifically trying not to measure science knowledge. The subtests are related: in 2025 scaled scores correlated .66 between Decision Making and Verbal Reasoning, .59 between Quantitative and Verbal Reasoning and .75 between Quantitative Reasoning and Decision Making, which the report reads as a set of common qualities measured across the subtests. It does not call that IQ. We found no published correlation between UCAT scores and an IQ test, and no conversion.

The usual way of putting a score on the IQ scale would mislead here anyway. The decile table is very close to a normal curve: 2100 is 0.85 standard deviations above the 2025 mean, the 80th percentile on a normal curve and exactly the Consortium's 8th decile, so the arithmetic in our z-score guide works. But the reference group is 41,354 people who chose to sit a test for medical and dental school entry, a self-selected group that will not resemble the population. A 90th percentile among them is not a 90th percentile in the population, so calling it an IQ of about 119, on the scale with a mean of 100 and a standard deviation of 15, would be wrong by an amount nobody can state. Our percentile calculator is for scores whose norm group is the population.

How well does the UCAT predict success in medical school?

Modestly. McManus and colleagues followed 4,811 students at 12 UK medical schools who took the UKCAT in 2006 to 2008 and found a correlation of .148 between the UKCAT total and first-year overall marks. Prior educational attainment correlated .362 with the same marks in 3,432 students. Squared, .148 is about 2 per cent of the variation in first-year marks and .362 about 13 per cent, our arithmetic. The authors describe the added value of the UKCAT over educational attainment as significant but small.

A later study of 6,812 entrants (Tiffin and colleagues, BMC Medicine, 2016) reached a similar verdict of statistically significant but modest incremental predictive validity, and a 2022 systematic review of 24 studies (Bala and colleagues, Medical Teacher) found that for over 70 per cent of univariate data points the UCAT had no statistically significant predictive validity and that for the rest its power was weak. Those studies cover the earlier test, with Abstract Reasoning, which the Consortium dropped after its own analysis found lower predictive validity than the other sections. Our guides to achievement tests versus IQ tests and to whether IQ predicts school success set this in context.

Does coaching help, and who gets it?

The Consortium calls commercial preparation unnecessary and unregulated, says it does not automatically equate to better performance, and advises scepticism about coaching claims. It gave coachability as a reason for removing Abstract Reasoning, and its technical report says scores on that subtest rose over a decade while response times fell. Observational evidence points to a gap in who prepares. Kulkarni, Parry and Sitch (BMC Medical Education, 2022) questioned UK school-leavers who sat the 2017 test: 5,439 of 14,332 replied, 38 per cent. Using the official timed practice tests was associated with the largest gain, 67.77 points (95 per cent interval 52.01 to 83.53) on the older 1,200 to 3,600 scale, and paid commercial materials with 37.65 points. On that scale 80 points was the step from the 5th to the 6th decile. Candidates from less deprived backgrounds and from independent or grammar schools were significantly more likely to use paid materials and to prepare for longer. The data are self-reported and observational, so they show association, not cause.

Any timed reasoning test rewards practice, as our guide to the practice effect shows. Other entrance tests raise the same questions about what a score means: see our guides to JEE and IQ and to the digital adaptive SAT, and the way scores describe a candidate relative to a cohort in our guides to 11-plus scores and CAT4 scores.

A UCAT score tells a medical school where you stand among that year's applicants. The IQ Metrics test reports a score on the standard IQ scale with its percentile, and our score converter translates between published scales when a test has the norms to allow it. You can try the underlying reasoning skills in our verbal and numerical IQ tests.

Common questions