Mathematics and Computer Science at Oxford

Oxford ranks #1 in the available Guardian and Complete University Guide Computer Science ranking display used for this page, but the caveat matters: those rankings are for Computer Science / Computer Science and Information Systems, not specifically this joint course.

The comparison set places Oxford #1 in both available columns, with Cambridge #2, Imperial #4 Guardian / #3 Complete, St Andrews #3 Guardian / #4 Complete, and Birmingham #5 in both columns. Use those rankings as a broad context signal rather than as a substitute for course fit.

The substantive reason to consider Oxford here is the combination of proof-based mathematics, theoretical computer science, algorithms and practical programming across the same degree route. If you want a course where abstract reasoning and computational implementation develop side by side, this is the right kind of structure to examine closely.

In our view, Oxford is the stronger fit if you want a collegiate course with intensive academic interviews and a proof-heavy joint structure. Imperial may suit applicants who prefer a London-based technology-campus environment, while Cambridge may suit applicants who prefer its course architecture and college teaching pattern.

Oxford requires the Test of Mathematics for University Admission (TMUA) for Mathematics and Computer Science in the 2027 entry cycle. TMUA is delivered by UAT-UK through Pearson professional test centres.

The required TMUA papers are Paper 1: Applications of Mathematical Knowledge and Paper 2: Mathematical Reasoning. The test window is 12–16 October 2026, with registration opening on 1 June 2026 at 3pm BST and closing on 28 September 2026 at 6pm BST.

This is a cycle change: from the 2027-entry cycle, with the test taken in October 2026, TMUA replaces MAT for this route. For applicants who have older advice saved, this is the detail to update first.

TMUA is a material part of shortlisting, and Oxford uses course selection criteria and the TMUA result to decide whom to invite for interview. UAT-UK indicates there is no pass/fail score, but Oxford does not publish a score threshold or fixed weighting.

For international applicants, TMUA gives Oxford another way to compare applicants taking different qualifications. Candidates testing in China, Hong Kong or Macau have a restricted 15–16 October window within the October sitting.

We recommend starting with the official TMUA specification, then moving into timed papers and error review. A useful next page is our [TMUA guide](/admissions-tests/tmua/).

For this course, the personal statement should show how your interest in mathematics and computer science connects. Avoid a list of technologies; Oxford’s course is built around proof, algorithms, models of computation and programming foundations.

A useful paragraph often starts with one problem, theorem, algorithm or project. Explain what you first thought, what changed your mind, and what you learned about reasoning.

We recommend using one mathematics example and one computing example, then adding one bridge between them. For example, an algorithmic puzzle can become interesting because of its proof of correctness, not just because the code runs.

Do not over-claim. It is better to explain a small project precisely than to describe artificial intelligence, cryptography or quantum computing in slogans.

Start with the Oxford Mathematics and Computer Science course page because it gives the official course identity, requirements and course outline. For the test, use UAT-UK TMUA information as the official test resource.

For subject-building, MIT Mathematics for Computer Science is a useful open-access archival course because it links proof, discrete mathematics and computational thinking. Project Euler is useful if you treat each problem as a reasoning exercise, not a race to code.

For admissions context, the Oxford Department of Computer Science undergraduate admissions statistics help you understand the Computer Science-family process. International applicants should also keep the Oxford international qualifications page close while checking qualification equivalence.

The British Informatics Olympiad is worth using as algorithmic practice if you review why a solution works, not only whether it passes tests.

30 colleges offer this subject. 22.2% of applicants submit an open application. 35.8% of places come through the pool.

Oxford is collegiate for this course. The exact 39-college count was not verified in the checked undergraduate college-choice page, so this draft does not use that figure.

Around a fifth of applicants make open applications; the 2023-24 Computer Science-family report recorded 22.2% open applications, but this is a CS-family figure rather than an MCS-only statistic. Around a third of successful applicants receive an offer from a college they did not specify; the same CS-family report recorded 35.8% offers by non-applied colleges, again with the CS-family caveat.

Oxford calls this process reallocation. It is used to even out competition, assign open applications to a college or hall with relatively fewer applications, and allow applicants to be interviewed by another college when a college is oversubscribed.

College choice affects tutorials, accommodation and social base, but it is not a tactical shortcut because colleges use a shared admissions framework. We recommend choosing a college you would be happy to live and study in, or making an open application if you do not have a strong preference.

Oxford considers academic achievement in context where possible. Contextual processes can support applicants from disadvantaged or disrupted backgrounds.

For this course, context can include whether Further Mathematics was available at school, because Oxford lists alternative routes where Further Mathematics is unavailable. Mitigating circumstances should be communicated early through UCAS, the reference or the relevant Oxford process.

We recommend making the context specific and evidenced. A clear sentence from a referee about subject availability or disruption is more useful than a vague explanation added late.