TMUA preparation guide

TMUA candidate numbers: how big the 2026 cohort really was

The size of the TMUA gets guessed wrong in both directions: some applicants treat it as a niche paper, others imagine a six-figure field. UAT-UK's 2025-26 annual report and the TMUA technical report settle it in print, and this page lays those figures out in the order a candidate actually needs them: how large TMUA is among the three UAT-UK tests, how bookings differ from tests actually sat, how the cohort splits across the two sittings and across UK and overseas, and what the official portrait of this year's candidates looks like. The last section covers the half that matters just as much, which is where these counts stop, and why none of them can be turned into an admissions figure for any university.
Practise 28 TMUA papers (2 free)Full TMUA guide
Papers
28
Questions
827
Free to try
2
Quick answer: how many people sit the TMUA
Tests delivered
16,092 delivered, from 16,888 bookings (Annual Report, Table 1)
Share of UAT-UK
48.9% of all tests taken, the largest of the three; ESAT 43.1%, TARA 8.0%
The two sittings
9,685 in October 2025 (60%) and 6,407 in January 2026 (40%)
Can you derive an acceptance rate
No. The reports carry candidate data only, with no admissions figures from any university

Every figure here comes from a printed table or sentence in UAT-UK's 2025-26 annual report and TMUA technical report. Nothing is read off a chart image and nothing is added or subtracted across tables. FrontierVue is an independent preparation platform and is not affiliated with UAT-UK, Pearson or any university.

In short
  • TMUA is the largest of UAT-UK's three tests: 16,888 bookings and 16,092 delivered tests, 48.9% of everything sat that cycle, against 43.1% for ESAT and 8.0% for TARA.
  • Bookings and delivered tests are two separate figures printed side by side on every row, and the annual report footnotes the booking total to explain the gap: it is inclusive of no-shows, candidates who did not cancel but were not present, so no score is reported for them.
  • The two sittings: 9,685 candidates in October 2025 (60%) and 6,407 in January 2026 (40%). The technical report states why in so many words: a candidate may sit only once per cycle, and Cambridge applicants must sit the October event.
  • Year-on-year growth is very uneven between the events: the annual report prints TMUA delivered tests up 8% in October and up 32% in January against 2024-25. That is growth in headcount, and says nothing about comparing scores across cycles.
  • Neither report carries application or admissions figures from any university, so candidate counts cannot be turned into an acceptance rate, and the per-country bars exist only inside a chart image, so no official figure supports any claim about how many candidates sat TMUA in a given country.
01

Scale: 16,092 tests, the largest of UAT-UK's three

The annual report counts all three tests on one page: 34,475 candidates booked a UAT-UK test for 2026 entry, and TMUA alone accounts for 16,888 of those bookings and 16,092 delivered tests. On tests actually sat, TMUA is 48.9% of the total, the largest of the three. Note that the report prints bookings and delivered tests as separate columns with separate shares; the table below keeps that layout.

16,092

TMUA tests actually delivered in the 2026 admissions cycle

48.9%

TMUA's share of every UAT-UK test taken, the largest of the three

16,888

TMUA bookings, printed as its own column and larger than the delivered figure

34,475

Candidates booking a UAT-UK test for 2026 entry, all three tests combined

ESATTARATMUA
Total booked14,781280616,888
Share of bookings42.9%8.1%49.0%
Total delivered14,193262616,092
Share of tests taken43.1%8.0%48.9%

Annual Report, Table 1, as printed. The ESAT and TARA columns are here for comparison only; each has its own page on this site.

02

Bookings and sittings are two different numbers

Whole cycle, booked

16,888

The Total Booked column of Annual Report Table 1

Whole cycle, delivered

16,092

Total Delivered in the same table, and the N behind the technical report's score tables

October 2025

10,109 booked, 9,685 sat

TMUA was 46.7% of everything booked in that window

January 2026

6780 booked, 6,407 sat

TMUA was 52.9% of everything booked in that window

Every row shows more bookings than delivered tests, and the report does not leave that gap open: the footnote hanging off the booking total says it is inclusive of no-shows, candidates who did not cancel their booking but were not present for the test, so no score is reported for them. This page quotes that footnote and stops there, because the report itself prints no rate derived from it. Neither figure, meanwhile, is an application count for any institution, since these documents carry no admissions data. The real hazard is mixing the two, for instance treating the booking figure as the field you competed against while reading percentiles computed on the delivered figure.
The working rule is simple. To place yourself on the score distribution, use 16,092: Table 4 of the technical report, whose overall mean is 4.28, is computed on exactly that population. To describe demand for the test, use 16,888. If both appear in one sentence, say which is which.
03

When and where: sittings and geography

October 2025January 2026
TMUA bookings10,1096780
TMUA tests delivered9,6856,407
Change in delivered tests from 2024-25+8%+32%
TMUA's share of that window's bookings46.7%52.9%
TMUA bookings inside the UK56843897
TMUA bookings overseas44252882

The first four rows come from Annual Report Table 2, the last two from Table 3. Table 3 prints bookings only, so there is no delivered-test split between UK and overseas.

You do not have to guess why one window is bigger, because the technical report says it outright: a candidate may sit only once within an admissions cycle, and those applying to the University of Cambridge must sit the October event, which is what accounts for the higher volume there. A quieter contrast sits in the same table. TMUA made up 46.7% of October's bookings but 52.9% of January's, so TMUA was the majority of the January window rather than a share of it. Choosing between the two is a different question, handled on October versus January with the per-sitting score distributions.
The centre network sets the boundary of where this cohort could sit at all. UAT-UK uses Pearson as its delivery partner for its network of over 5500 professional test centres across 180 or more countries and territories, and in 2025-26 tests were actually taken in 129 different countries. The key markets outside the UK named in the report were similar to the previous year: China, India, Singapore, Malaysia, Hong Kong, the United States, the UAE, Türkiye, Spain and France. Read that list as covering all three tests together, not as a TMUA-only breakdown.
04

Who sat it: the official portrait of the cohort

Gender
11,253 candidates identified as man (70%) and 4,642 as woman (29%). The technical report prints only those two rows, and the two percentages total 99%, with no further category printed as text.
Area of residence
UK 8,643 candidates (54%), Other 6,733 (42%), EU 716 (4%). The question asked is country of permanent residence, not where the test was sat. The report stops at these three categories and goes no further down towards individual countries.
First language
9,255 candidates gave English as their first language (58%) and 6,837 gave another language (42%). The previous cycle stood at 60% English, so the non-English share is rising. Note that this 42% and the 42% under residence are numerically equal but are not the same people.
School type, UK candidates only
Further Education or Sixth Form College 3478 (40%), selective grammar school 1887 (22%), independent or private fee-paying school 1616 (19%), academy or secondary school 1506 (17%), university 86 (1%) and other 68 (1%). This breakdown covers UK candidates only; overseas candidates are not included.
Background and health
Among UK candidates, 700 were in receipt of free school meals (8%) and 7,364 were not (85%). Across the whole cohort, 12,604 candidates said a parent or guardian had attended tertiary education (78%) and 2,178 said they had not (14%). On learning difficulty or chronic health condition, 932 answered yes (6%) and 14,222 answered no (88%).
Bursaries and access arrangements
The bursary table is broken down by sitting, not by test: 2379 issued across the year, 12.8% of UK candidates, of which 1300 in October (11.2%) and 1079 in January (15.4%), with 94.1% of applications approved. Access arrangements do get split by test, where the TMUA row reads 676 and 4.2%, though that table is titled as delivered tests while its column header reads candidates booked. The technical report counts the same thing as 675 TMUA candidates, around 4%, of whom 664 asked for extra time and/or pause-the-clock.
This portrait answers who turned up for the TMUA, not who studies mathematics. Everyone counted here has already been filtered by the courses they applied to, so read it as a description of this test's field rather than a sample of any wider population. Two rows repay attention if you are outside the UK. On residence, 54% were resident in the UK and 42% outside both the UK and the EU. On language, 58% gave English as a first language and 42% did not. How those same groups differ on score is deliberately absent here and belongs to the two pages linked below.
Everything above is headcount and share, not performance. The score side of the same groups lives elsewhere: the whole-cohort distribution and group means are on TMUA score statistics, and the residence and first-language breakdowns are on TMUA international candidates. If you would rather build the skill before reading the tables, start from TMUA past papers and practice.
05

What these numbers cannot tell you

0134,475

Everyone who booked

For 2026 entry, 34,475 candidates booked a UAT-UK test across the three papers. This is the largest number in the chain.

0216,888

Narrowed to TMUA

Of those, 16,888 bookings were for the TMUA, 49.0% of all bookings.

0316,092

Actually sat

16,092 TMUA tests were delivered. Every score statistic in the technical report is computed on this population.

04No next step

The data stops here

Neither official report contains application volumes, offers made or places filled at any institution. The chain ends here, and any ratio extended past this point is invented.

Candidate counts are not an acceptance rate
Candidate numbers and places are different quantities. Six institutions used these tests for 2026 entry, but the reports list no application volume and no admissions outcome for any of them, so the division people want to perform has no denominator. This is not an acceptance rate and cannot be dressed up as one. On the same evidence there is no pass mark anywhere in these documents, and no score threshold of any kind is printed in them.
Headcounts compare across cycles; scores do not
The annual report prints the year-on-year change itself, up 8% in October and up 32% in January, and that is UAT-UK's own headcount comparison, safe to quote. The same body of documents is equally clear that the scaling constants are fixed within a single admissions cycle, so absolute scaled scores are not comparable between cycles. Growth in numbers and the meaning of a score are two separate lines; do not merge them into one sentence.
There is no per-country TMUA figure
The only per-country breakdown in the annual report is a bar chart whose values appear nowhere in the text layer, so this page prints none and you should be wary of any figure someone has eyeballed off that image. The report does carry a continent table, but it counts overseas candidates across all three tests together and cannot be read as a TMUA-only distribution.
The official documents print more than one total
Quote the table number along with the figure. Annual Report Table 1 and technical report Table 4 both give 16,092, while the technical report's discussion of unreached items refers to 16,091 candidates; access arrangements appear once as 676 tests and once as 675 candidates. The differences are small and change nothing, but both values genuinely sit in the official documents, and citing the source beats quietly picking one.
Treat this page as a factual base rather than a conclusion. It answers one class of question: how large the test is, roughly who takes it, and how big the pool was that produced your percentile. What score you personally need depends on the course and the institution and how they used the test that year. Neither report answers that, and this site will not invent a number on their behalf.
FAQ

Frequently asked questions

How many people sat the TMUA for 2026 entry?
Annual Report Table 1 gives two figures: 16,888 bookings and 16,092 delivered tests. The technical report's score statistics are computed on 16,092. On tests taken, TMUA was 48.9% of everything UAT-UK delivered. Split by event, 9,685 sat in October 2025 and 6,407 in January 2026.
Is TMUA the largest of the three UAT-UK tests?
Yes. On tests taken it is 48.9% for TMUA, 43.1% for ESAT and 8.0% for TARA; on bookings it is 49.0%, 42.9% and 8.1%. TMUA leads on both measures, though not by much over ESAT.
Why was the October sitting so much bigger?
The technical report gives the reason in plain terms: a candidate may sit only once within an admissions cycle, and applicants to the University of Cambridge must sit the October event, which accounts for the higher volume there. Taking all three tests together, 63.1% of the cycle's tests were delivered between 12 and 16 October 2025. The share runs the other way, though: TMUA was 46.7% of October's bookings but 52.9% of January's.
How many candidates sat the TMUA in mainland China?
No official document prints it. The only per-country breakdown in the annual report is a bar chart whose values never appear as text, and the continent table counts overseas candidates across all three tests, so it cannot stand in for a TMUA figure. What can be said: tests were taken in 129 countries in 2025-26, China is one of the key non-UK markets the report names, and 6,733 TMUA candidates (42%) were resident outside both the UK and the EU.
More people are sitting it. Does that mean tougher competition?
Numbers are indeed rising: the annual report shows TMUA delivered tests up 8% in October and up 32% in January against 2024-25. That tells you more people sat the test, and nothing else. The reports carry no application volume, no offers made and no places filled for any institution, so an acceptance rate cannot be computed and there is no way to tell whether places moved in step. Candidate counts are one half of that fraction; the other half is not in these documents.
SOURCES

Official sources

MORE GUIDES

Other TMUA preparation guides

FrontierVUE is an independent practice platform. It is not affiliated with or endorsed by UAT-UK, Pearson, OCR, the University of Cambridge, Imperial College London, or any official admissions-test owner.

FrontierVUE 是独立的备考练习平台，与 UAT-UK、Pearson、OCR、剑桥大学、 帝国理工学院或任何官方入学考试主办方均无隶属或背书关系。

Put it into practice

28 TMUA papers on FrontierVUE, including 10 Frontier Original mocks built to the current format, every question with a worked solution. The first 2 papers are free.

Practise TMUA papersSee pricing