Important update: The Oxford Maths Admissions Test (MAT) has been discontinued.
The MAT was used by the University of Oxford from 2007 to 2025. In 2026, Oxford replaced the MAT with the Test of Mathematics for University Admission (TMUA).
All applicants for Mathematics, Computer Science and relevant joint honours courses at Oxford must now take the TMUA instead.
Applying to Oxford to study Maths? Read our Guide to the TMUA for the latest information on the test, format, preparation and key dates.
What was the MAT?
The Mathematics Admissions Test (MAT) was a university admissions test designed to assess mathematical ability and problem-solving skills.
Oxford used the MAT for applicants to Mathematics, Computer Science and other related courses. The test was introduced in 2007 and was used until 2025.
Applicants for Mathematics, Computer Science and relevant joint honours courses now take the Test of Mathematics for University Admission (TMUA) instead.
The change means that students applying to Oxford for Mathematics or Computer Science should no longer prepare for the MAT as their admissions test.
What replaced the MAT?
The TMUA (Test of Mathematics for University Admission) replaced the MAT at Oxford from the 2026 admissions round.
For 2027 entry, Oxford requires the TMUA for:
- Mathematics
- Mathematics and Statistics
- Mathematics and Computer Science
- Mathematics and Philosophy
- Computer Science
- Computer Science and Philosophy
Applicants for these courses must take both papers of the TMUA.
The TMUA is designed to assess mathematical reasoning and the ability to apply mathematical knowledge to unfamiliar problems. There is some continuity between the skills required for the old MAT and the new TMUA, even though the tests have different formats. Oxford notes that the mathematical topics covered by the two tests are very similar and that both emphasise problem-solving.
What did the MAT test?
Although the MAT has been discontinued, its approach is still useful to understand because it explains why problem-solving ability was so important in Oxford's mathematics admissions process.
The MAT was not simply a test of whether students had memorised the material covered in their A-level Maths course. Instead, it was designed to assess how students approached unfamiliar mathematical problems.
How was the MAT structured?
The MAT lasted 2½ hours and consisted of 7 questions. Each candidate answered 5 of these questions. The questions that you were required to answer depended on your degree choice. For instance, Maths candidates had to answer questions 1-5, but Computer Science or Maths with Computer Science students will each have a different selection of questions.
Question 1, which all students needed to attempt, was composed of ten multiple choice questions (with five answer options) worth 4 marks each, for a total of 40. The remaining four questions were longer questions and were usually composed of many related parts. Each full question was worth 15 marks each, for a total of 60.
What topics did the MAT cover?
The MAT syllabus was based primarily on the pure mathematics content of AS-level Maths, with some additional topics.
The test was deliberately designed so that mathematical difficulty did not simply depend on how advanced the mathematical content was. A relatively elementary mathematical idea could be used to create a demanding problem. This is one reason why MAT preparation focused so heavily on problem-solving rather than simply learning additional mathematical content.
The TMUA has a different specification, so current applicants should use the official TMUA specification rather than relying on the old MAT syllabus.
What skills did the MAT assess?
The goal of the MAT wasn't to test that you knew the AS syllabus, it was to see how you responded to questions with a large problem-solving component. That is, a question which is difficult not because of the fact that it explicitly involves complex topics, but because the process of navigating yourself to a solution is complex: it relies on a stroke of inspiration, backed up with a systematic and logical approach, in order to solve it.
Taking this idea of problem solving vs topic complexity to its extreme, you get questions like the following one from Round 2 of the British Mathematical Olympiad (BMO) in 1994:
“Find the first integer n, which is greater than 1 and is such that the average of 12, 22, 32, ... , n2 is itself a square number.”
You could ask an 11-year-old this question and they would probably be able to understand it, and yet it appears on perhaps the most difficult Maths test that someone could sit when they are 18 years old. The difficulty arises due to the fact that at the beginning, there seems to be no clear path that will definitely solve the problem (in a reasonable amount of time). A question like this illustrates one of the central challenges of mathematical problem-solving: knowing how to proceed when there is no immediately obvious route to a solution. The best strategy is often to think of a few ideas and then get stuck into the most appealing one as soon as possible - you will never get anywhere with a blank page. Here are a few strategies that often work, taken from nrich’s guide to problem solving:
- Try special cases or a simpler problem.
- Work backwards.
- Guess and check.
- Be systematic.
- Work towards subgoals.
- Imagine your way through the problem.
- Has the plan failed? Know when it’s time to abandon the plan and move on.
In the example above, you might try special cases e.g. averaging the first few squares to see if you can see a pattern emerging. You could also try to be systematic by writing a general equation which you then try to solve. However, you should keep in mind that at the beginning of a problem like this that you are only experimenting. You are not necessarily hoping to solve the problem from the outset: you only want to gain an insight into how the problem should be solved. As such, you should be willing to abandon a solution that doesn’t look like it will lead anywhere, even if you have invested 5 or 10 minutes into it already. Often, you can still use what you have tried to inform your next attempt.
Is the TMUA harder than the MAT?
It is difficult to make a direct comparison because the tests have different formats. The MAT included longer written questions, while the TMUA consists of two multiple-choice papers. Both are/were designed to assess mathematical thinking and problem-solving rather than simply recall of mathematical facts.
How should I prepare for Oxford Mathematics?
Preparation for a competitive mathematics degree should ideally begin well before the admissions test.
Students should first make sure they have a strong understanding of the relevant mathematical material. They can then focus increasingly on unfamiliar problems that require them to apply that knowledge.
MAT past papers can still be useful for developing mathematical problem-solving skills, but current applicants should prioritise TMUA past papers and official TMUA preparation materials. The two tests have different formats, so MAT papers should be treated as supplementary practice rather than a substitute for TMUA preparation.
There are also several useful mathematical problem-solving resources, including some resources originally developed for MAT preparation:
- Use Underground Mathematics' bank of MAT questions for additional problem-solving practice. These questions were designed for the MAT, so they should be treated as supplementary practice rather than a substitute for TMUA preparation.
- Check out nrich’s Advanced Problem Solving Modules for a well-crafted series of questions to guide you through the process of learning to problem solve in a mathematical context.
- Have a look at Cambridge’s STEP support Programme for another guided approach to learning to problem solve. This page also contains a link to Dr Stephen Siklos’ book on STEP preparation, which is an excellent (and free) resource.
Oxford TMUA tutors
Keystone Tutors represents specialist TMUA tutors who can provide personalised preparation for students applying to competitive Mathematics and Computer Science courses. Find out more about TMUA tutoring.