Imperial TMUA: 19 courses, and the score data Imperial publishes

Imperial requires the TMUA for 19 undergraduate courses. Every one of them is compulsory.

Imperial also uses the ESAT, for around forty other courses. The heading on the Imperial section of the official list reads "either the ESAT or the TMUA is compulsory", and which one you sit depends entirely on your course code.

Here's how that breaks down:

That last row is the one most pages miss. BSc Economics, Finance and Data Science (L1N3) sits in the Business School rather than in Computing or Mathematics, and it is just as compulsory as the other eighteen.

The full seven-university list is in our guide to which universities require the TMUA.

Imperial publishes an interactive dashboard of TMUA scores from the 2025 entry cycle, broken down by department, by fee category and by application outcome. It is anonymised, it is primary data, and it is free to read.

It is also awkward to read. The figures live inside an interactive report with a department filter, four departments and three fee categories, and nothing on the page pulls them together into a single view.

Every table in this section is our own analysis: we worked through each department and each fee category in turn, recorded the figures on 3 September 2026, and laid them out side by side so you can compare them.

The dashboard itself shows you one slice at a time.

Starting with the headline comparison: what offer holders scored, against what everyone else scored.

Here's how that breaks down:

Three things jump out.

1. The gap between offer and no offer is about 3 points, everywhere. That is large, and it is consistent across four departments with very different intakes.

2. There is no single Imperial number. The Maths offer mean of 7.6 and the Business School offer mean of 5.3 are more than two points apart. Anyone quoting one figure for Imperial is quoting the wrong one for three quarters of readers.

3. The Business School offer mean, 5.3, is barely above the Maths rejection mean of 4.7. A score that would be unremarkable in the Maths pool is above average in the Business School pool.

The same data split by fee category

Imperial splits every figure three ways:

Home non-WP, Home WP (widening participation) and Overseas. Reading the score alongside the offer rate for each group is where the dashboard earns its keep, so we have put both in one table.

Here's how that breaks down:

Take the Mathematics row. Offer holders in the widening participation group averaged 5.7, and overseas offer holders averaged 8.4.

That is a spread of 2.7 points for the same offer, on the same course, in the same year.

Then look at what happened to those two groups. Home WP applicants had the highest offer rate of the three at 24.1%, while overseas applicants had the lowest at 16.6%, despite scoring 2.7 points higher on average.

How the dashboard maps onto the 19 courses

The dashboard has four departments and they account for all 19 courses, so you can find your own.

The dashboard also carries a stacked bar chart of applicants across the score range and a box and whisker plot of the distribution by outcome. Those are worth looking at for the shape of the distribution, which the means alone do not give you.

The proportion of applicants receiving an offer climbs steeply with score, from close to zero at the bottom of the range to the large majority at the top.

One thing the dashboard does not give you is numbers of applicants. Every figure is a percentage or a mean, so there is no published count of how many people applied to each department or sat the test for it.

We have not estimated one.

The single most striking number Imperial publishes is 8.4: the mean TMUA score of overseas applicants who received an offer for Mathematics in the 2025 entry cycle.

For scale, the 90th percentile of the entire TMUA cohort for 2026 entry was 6.7, and the scale caps at 9.0. An average of 8.4 puts the typical overseas Maths offer holder in the top few per cent of everyone who sat the test that year.

The pattern holds across every department.

In every one of the four departments, overseas offer holders scored higher than Home non-WP offer holders. In three of the four, they were also less likely to receive an offer.

The Business School is the sharpest case: overseas applicants averaged more than a point higher and were offered a place at less than half the rate.

Imperial does not explain why, and we are not going to guess at admissions policy. What can be said is what the numbers describe.

1. The overseas pool is more competitive. Whatever the cause, an overseas applicant is being compared against a group with a higher average score and a lower rate of offers.

2. A score that is strong in one pool may be ordinary in another. A 7.0 was the Home non-WP average for a Maths offer. In the overseas group the average was 8.4.

3. The gap is largest where the course is smallest. Mathematics and the Business School show the widest overseas-to-home differences of the four.

One practical consequence worth stating: the fee status Imperial uses here follows your fee category for tuition purposes, not your nationality or where you sit the test.

The TMUA fee works differently again, at £78 in the UK and Republic of Ireland and £133 elsewhere, set by test centre location rather than by nationality. Bursaries are available to UK candidates in financial need only, with no overseas equivalent.

Details are in TMUA dates, registration and booking.

Imperial say this in as many words:

"There aren't any grade boundaries or pass marks for the ESAT or TMUA."

And, on how the score is weighed:

"Achieving a high score does not necessarily mean that you will be offered a place, while a low score does not mean that your application will be unsuccessful."

Imperial attach a warning to the dashboard as well, which is worth repeating in full because it is the correct way to read everything in the section above:

"This is historic data. Admissions test scores, score distributions, and offer outcomes can vary from year to year, and the scores shown here should not be taken as indicative of requirements or outcomes for the current or future admissions cycles."

The reason there are no boundaries is structural rather than a matter of policy. Scores are produced by post-equating across multiple test forms after the window closes, and the scale is anchored to the cohort.

Our guide to TMUA scoring and results explains the mechanism.

It is easy to turn a distribution into a target. Four reasons not to.

1. A mean is not a cut-off. The Mathematics offer mean of 7.6 means roughly half of successful applicants scored below it. It is the centre of a range, not its floor.

2. Scores a few tenths apart are not meaningfully different. The raw score standard error of measurement on the TMUA is 2.58 to 2.71 marks out of 40. Two candidates whose scaled scores differ slightly are, statistically, the same candidate. Treating 7.4 and 7.6 as different outcomes is reading noise.

3. These describe the applicants who applied rather than a standard Imperial set. The Maths pool self-selects heavily, so a high mean partly reflects who chose to apply.

4. It is one cycle, and Imperial say it may not repeat. 2025 entry data cannot be a requirement for 2027 entry, and Imperial explicitly decline to make it one.

It is worth knowing how unusual this data is. Imperial is the only TMUA university that publishes score distributions at all.

Warwick publishes a single figure, which describes offer holders on one course group rather than a distribution, and Durham publishes a threshold that converts into a grade reduction. Cambridge, Oxford, LSE and UCL publish nothing at all.

The useful way to use these numbers is as orientation rather than as a bar.

They tell a Business School applicant that a 5.3 was a normal offer-holder score last cycle, and a Maths applicant that the same 5.3 would have sat below the Maths average for applicants who were turned down.

That is genuinely worth knowing before you decide how much preparation the test deserves.

For the wider question of what counts as a good score across the whole cohort, and which other universities publish anything, see what is a good TMUA score.

Imperial applicants may sit in either window. There is no Imperial rule forcing October, unlike Oxford and Cambridge.

One caution on the January window. Across two consecutive cycles the January cohort has scored roughly 0.9 points lower on an October-anchored scale.

The cause is not established and the January group self-selects, so treat it as a pattern worth knowing rather than a sign that the January paper is harder. Worth weighing before you assume the extra term is free.

The full date chain for both windows, including access arrangements and bursary deadlines, is in TMUA dates, registration and booking.

Your score reaches Imperial on its own. UAT-UK and Pearson VUE match it to your UCAS application course by course, so there is nothing to send, no way to withhold it from one university and share it with another, and no appeal against the result.

Nineteen courses across three very different departments means the honest answer to "how much preparation do I need" is not the same for every reader.

1. Work out which pool you are in first. The four department rows above differ by more than two points at the mean. Find your course in the mapping table, then look at that department rather than at Imperial as a whole.

2. Do not set a target score. The measurement error alone is wider than most of the gaps people agonise over, and Imperial say a high score does not secure a place. Aim to be comfortably inside the distribution for your department rather than at a specific number.

3. Spend the time on Paper 2. About 38% of Paper 2 tests reasoning that no A-level maths course formally teaches: the language of logic, converse and contrapositive, necessary and sufficient conditions, quantifier negation and spotting errors in proofs. A-level covers the methods of proof but not that vocabulary, which makes it the part of the test where preparation moves you most.

The best free starting point is the archive. The official TMUA past papers from 2016 to 2023 are free and every released paper comes with worked answers from UAT-UK.

That is eight sittings and 320 questions, and it is the same content specification as the current test.

Our guide to Paper 2 covers the logic and proof material, and the specification page shows which topics recur most often. If you are starting from scratch, the complete TMUA guide is the place to begin.