Entry CriteriaUK and Ireland, medicine, GEM and dentistry

UCAT decile tables by test cycle

The test cycle is the year the test was sat. It feeds the following year of entry: the 2024 cycle is 2025 entry. Schools usually quote the entry year, so a cut-off in a 2025-entry policy is a 2024 cycle figure. Abstract Reasoning was removed from the 2025 cycle, taking the total from 1200-3600 down to 900-2700. The Consortium has restated the 2019-2024 cohorts without Abstract Reasoning, which means a historical cut-off can be converted inside its own cohort instead of being mapped onto a different year. That is what the converter does.

Four subtests, total out of 3600 (cycles up to and including 2024)

Blue cells are published by the UCAT Consortium. Cream cells are our own estimate, and here is exactly how they were made. The Consortium has taken down the statistics pages for 2019 to 2022, so the decile table for those years no longer exists anywhere. What does survive is the average total score for each of those years. Scores on a test like this spread out around the average in a bell shape, so if you know the average and how wide the spread is you can work out where each tenth of candidates falls. That is what was done, year by year. The method was then tested on 2023 and 2024, where the real table does still exist, and it never missed a decile by more than 11 points. So treat cream figures as good to about plus or minus 10 points, not as exact. One caveat worth knowing: the workings were not kept, so the fit cannot now be checked line by line. If anyone has screenshots of the original 2019 to 2022 pages, those would replace this outright.

Three subtests, total out of 2700 (the current scale)

ALL OF THIS TABLE IS PUBLISHED, including 2019 to 2022. The 2019 to 2024 columns are the Consortium restatement of those cohorts with Abstract Reasoning removed, so they are the same candidates measured the modern way, not a model. This is the table to quote to students. The estimated figures in the four-subtest table above are only needed to place an old raw score, which cannot be done from this table because such scores do not exist on the 2700 scale. The AMBER column is the 2026 interim table published to 13 September 2026, covering 20,794 candidates against 41,354 for the whole of the 2025 cycle. It is not comparable with the blue columns and will fall before the cycle closes.

How to use this

Take the cut-off, the cycle it belongs to, and find its position in the four-subtest table for that same cycle.

Read the equivalent figure off the three-subtest table for that same cycle. Same cohort, same year, so no cross-year drift.

Report the decile band rather than the converted score wherever you can. Bands are what stay stable between years; scores are only that year's currency.

Below the 1st or above the 9th decile there is nothing to interpolate between, so the converter returns a band rather than a number.

Never substitute UCAT ANZ decile tables. Those are Australian and New Zealand candidates and their means run 30 to 60 points higher.

What removing Abstract Reasoning did to the scale

The 2024 cohort measured both ways against the first cycle actually sat without Abstract Reasoning. Scores rose, and rose further the higher up the scale you look.

WHICH TABLE TO USE. Judge a current applicant against the 2025 column, not against a restated year. A score of 2010 is exactly the 7th decile in 2025 but sits at the 74th percentile on the restated 2024 table, so an older table flatters a candidate, and flatters them more the higher they score. Restated years are for placing a school historical cut-off. The live 2025 table is for placing a student. No cause should be read into the shift: a stronger cohort, preparation adapting to a three-section test, or less fatigue across a shorter exam would all look like this, and one year cannot separate them. Worth rechecking when the 2026 statistics are published, because if the rise continues then every school published threshold drifts down in decile terms without the school changing anything.

Interim tables versus final tables: the trap

The Consortium publishes a decile table during the testing window as well as at the end of it, and the two differ. Testing runs from July to late September and later testers score lower on average, so every decile drifts DOWN as the cycle closes. This has now been checked in two separate cycles and the pattern is the same in both.

THE RULE OF THUMB. Every decile falls by roughly 50 to 60 points between the middle of September and the close of the cycle, in both years checked, at every decile rank. That is stable enough to tell a student in September what their position will look like once the cycle closes, before the final table exists.

WHY IT MATTERS. The interim table is the pessimistic one, and it is the only one that exists when results come out and choices are being made. A candidate reading it under-rates themselves by roughly 5 to 9 percentile points, about half a decile. On the 2024 figures a score of 2500 looks like the 4th decile in mid-September and is the 47.5th percentile once the cycle closes; a 2580 looks like the median and is the 58.6th percentile. The error is largest in the middle of the range, where most applicants sit and where most shortlisting thresholds fall. Medical schools work from closed-cycle figures. Always check the date a table was valid to, prefer the end-of-cycle version, and where a school quotes a decile in its own admissions policy assume it means the final table unless it says otherwise.

Worked examples from the workbook

Four figures, two schools, four cycles, all landing in the 4th to 5th decile. That stability is why the band is the thing to quote.

Sources

https://www.ucat.ac.uk/results/test-statistics-2025/

https://www.ucat.ac.uk/results/test-statistics-2024/

UCAT Consortium test statistics, 2023 cycle, and the restated 2019 to 2024 decile and subtest tables, captured from the Consortium site

https://www.ucat.ac.uk/about-us/technical-reports/ - mean and standard deviation for the 2019 to 2022 cycles

The 2026 test cycle, for entry in September 2027

This is the cycle your current applicants have just sat, and the only table that exists for them. It is an INTERIM table, published to 13 September 2026 and covering 20,794 candidates, about half a full cycle. On the evidence of 2023 and 2024 an interim table always sits above the final one, because the candidates who test early score higher than those who test late. Treat every figure in the first column below as the pessimistic case and the projection as the realistic one.

HOW THE PROJECTION IS MADE, and how much to trust it. In 2023 and 2024 every decile fell by 50 to 60 points between the September table and the final one. Both of those years were on the 3600 scale, where the deciles span 780 points. On the 2700 scale they span 640, so the same effect is worth about 45 points. That is what has been subtracted. It is one step of arithmetic on two years of evidence, and no interim-to-final pair yet exists on the current scale, so treat the projection as a direction of travel rather than a number to quote. What it is NOT is a reason to tell a student their score will improve. Their score is fixed. What moves is where it sits against everyone else.

WHAT IT LOOKS LIKE SO FAR. Even after the projection, this cohort sits at or above last year at every decile: level at the 1st, then 40 to 60 points higher across the middle and upper bands. If that holds, a score sitting exactly on last year 7th decile, 2010, lands just past the 6th this year. It is a real movement but a small one, and it should not change anyone four choices on its own. The figure to repeat to students is the decile, not the score, because the decile is the thing that carries across years and the score is not.

Situational Judgement Test bands, 2026 interim

Interim, to 13 September 2026, 20,794 candidates. Band proportions are steadier year on year than the score deciles, so these move less between the interim and final tables.

https://www.ucat.ac.uk/results/test-statistics-2026/ - interim table, to 13 September 2026, read 16 September 2026

Means and subtest means, 2026 interim against 2025 final

READ THE MEAN THE SAME WAY AS THE DECILES. The Consortium publishes 1970 as a PRELIMINARY figure and says in terms that these scores should be treated with significant caution, because 20,794 tests is under half the cycle and roughly another 22,000 candidates are expected. The raw gap of +79 on last year is therefore not the real rise. Subtracting the same 45 point interim effect used for the deciles gives a projected final mean near 1925, about 34 points above 2025. That is consistent with what the deciles show, which is a mild check on both numbers. Note also that the mean sits almost exactly on the 5th decile, 1980, so the distribution is close to symmetrical and the mean and the median can be used interchangeably when talking to students.

THE SUBTESTS ARE WHERE IT GETS INTERESTING. All three are up, but not evenly: Quantitative Reasoning +38, Decision Making +29, Verbal Reasoning +12. Part of that is the interim effect again, but the RANKING is not affected by it, and the ranking is stable year on year. Verbal Reasoning is the lowest-scoring subtest by a wide margin, 614 against 699 for Quantitative Reasoning, and the gap has widened rather than closed. A candidate scoring 614 in Verbal is exactly average in it; the same 614 in Quantitative would be well below average. Students who report a weak Verbal score are usually describing a normal one.

SJT HAS MOVED THE OTHER WAY. Band 1 has fallen from 21 per cent to 16, band 2 risen from 39 to 42 and band 3 from 29 to 32. Band 4 is unchanged at 10 per cent. So the cognitive scores are up and the SJT bands are slightly down. These are separate instruments and there is no reason they should move together, but it is worth knowing before anyone reads a single narrative into the cycle.

How a decile is stated in this workbook

A decile is used two different ways and mixing them causes off-by-one errors. A decile BOUNDARY is a single score: the 5th decile is the score at the 50th percentile. A decile BAND is the tenth of candidates between two boundaries: the 6th band runs from the 50th to the 60th percentile. This workbook reports BANDS, written as "5th band", except where a figure lands exactly on a boundary, which is written as "exactly the 5th". A score sitting between the 5th and 6th boundaries is in the 6th band, not the 5th. Checked and corrected throughout on 16 September 2026.

Which figures in this workbook rest on an estimated table

Only conversions of raw scores from the 2019 to 2022 test cycles, which means entry years 2020 to 2023. In the workbook that is: Glasgow entry 2020, 2021, 2022 and 2023; Warwick entry 2022 and 2023; King's 2019, 2020 and 2023 entry; Manchester dentistry 2021 entry; Newcastle dentistry 2020 entry; Cardiff dentistry 2023 entry. Everything from the 2023 test cycle onwards, which is entry 2024 and later, rests on published tables at both ends. The figures students actually need are all in the recent group, so the estimated ones affect the history rather than the advice.