What Is a Good TMUA Score? Published Percentiles 2024-26

A good TMUA score is one above the middle of your sitting: in October 2025 the median was 4.5 and the 90th-percentile score 7.0, while in January 2026 they were 3.6 and 5.9. UAT-UK publishes no pass mark. The technical reports (Table 2) fix the scale so that the median October candidate scores 4.5 and the 90th percentile scores 7.0 (one paragraph of UAT-UK’s Test Results page prints 9.0 for the 90th percentile, which conflicts with the reports and with that page’s own other text), and January candidates are placed on the same scale. This page sets out every overall-cohort percentile mark in Table 4 of UAT-UK’s two technical reports (October 2024 to January 2026), lets you check your own score against them, and explains why a score from before 2024 cannot be read the same way. ProSyllabus is an independent study site, not affiliated with UAT-UK or any university. UAT-UK’s pages and documents were read on 23 October 2026; the university pages and UAT-UK’s past-paper listing on 18 October 2026.

UAT-UK TMUA Technical Report 2025-26, Table 4, read 23 October 2026.

UAT-UK does not keep a fixed conversion from answers to scores. After each October sitting it fixes the scale on the October candidates: their median ability becomes a scaled score of 4.5 and the ability at their 90th percentile becomes 7.0. A straight line through those two points turns every candidate’s ability estimate into a score, capped at 1.0 and 9.0. The technical report says “The same scaling constants were used for both the October 2025 and January 2026 events to ensure the scaling was consistent and scaled scores were comparable across events.”

So an October median of 4.5 is not a sign that a sitting was hard or easy. It is how the scale is defined. A 4.5 is defined relative to each cycle’s October candidates, so do not treat it as a fixed measure of attainment across cycles.

The January 2025 median was 3.3 and the January 2026 median 3.6, against 4.5 in both Octobers. Because January candidates are scored with the October constants, this is a real difference in performance on one scale, not a different scale. UAT-UK does not say why. Two facts it does state: most Oxford and Cambridge applicants must sit in October (its exceptions are mature applicants to a Cambridge mature college with a January deadline and applicants to Oxford’s Astrophoria Foundation Year programme), and, for applicants to all other institutions, “There is no advantage to sitting in either session, but you can only sit once within an admissions cycle.”

In practice: a 5.0 in January 2026 was between that sitting’s 75th-percentile score (4.6) and its 90th-percentile score (5.9), but between the median (4.5) and the 75th-percentile score (5.8) of the October 2025 candidates. Universities see one number on one scale, so compare yourself with the October figures as well as your own sitting.

No university checked for our guide to scores by university publishes a TMUA pass mark. Two publish figures you can measure against, both from their own pages:

Put next to the percentiles, Imperial’s Mathematics offer holders (mean 7.6) averaged above the October 90th-percentile score, and Warwick’s 5.0 line sits a little above the October median. For every course that requires the test and what each university says, read what TMUA score you need for Oxford, Cambridge, Imperial and LSE.

UAT-UK’s 2025-26 technical report gives a standard error of measurement of 0.62 to 0.64 for the October 2025 scaled scores. Read the way the report describes it, there is about a 68% chance that a candidate’s “true” score lies within one standard error of the score reported. Our reading, not the report’s: a 5.0 and a 5.4 are closer than they look; a 4.5 and a 7.0 are not.

UAT-UK says October 2026 candidates receive their results in their UAT-UK account on Monday 16 November 2026, and January 2027 candidates on 8 February 2027. It also uploads an explanation of results to its Test Results page. The technical reports for the last two cycles, which carry the percentile table, were published in September 2025 and July 2026; we will add the new figures if and when UAT-UK publishes them.

UAT-UK recommends answering practice questions under timed conditions. Its 18 official papers (Paper 1 and Paper 2 for 2016–2023, plus an early specimen pair) are free (see our TMUA past papers guide); our TMUA practice tests add two full mocks in the same format, 20 questions in 75 minutes per paper, with the working for every answer. Try the five free questions first. Here is one of them: