UK Admissions Tests & the Oxbridge Interview
The 2026 overhaul most preparation advice hasn't caught up with — and what the interview really is
The 2026 overhaul most preparation advice hasn't caught up with — and what the interview really is
An Empower Education resource. Written for Hong Kong students and parents. Last reviewed for the 2026 admissions cycle — always confirm current requirements directly with UAT-UK, UCAT, LNAT, and each university.
If your family has been given a reading list for the MAT, a tutor for the PAT, or past papers for the TSA, you have been prepared for tests that no longer exist.
From 2026 entry, Oxford retired almost its entire suite of in-house admissions tests — the MAT, PAT, TSA, CAT, MLAT, PhilAT, BMSAT and AHCAAT among them — and replaced them with three computer-based tests run by UAT-UK, a joint venture between Imperial College London and the University of Cambridge. Only the UCAT for Medicine and the LNAT for Law survive unchanged.
This is the largest change to UK admissions testing in a decade, and a great deal of tutoring, guidance, and parent-group wisdom in Hong Kong has not caught up with it. Families are still buying preparation for retired exams.
This guide covers what the tests actually are now, the registration trap that catches international families every year, and what the Oxbridge interview genuinely assesses — using Oxford's own published questions rather than the folklore.
The one idea
The tests are shortlisting instruments, not knowledge exams. They exist because too many applicants have perfect grades — so they measure how you think when the syllabus runs out.
This is why cramming content works poorly, and why a student who has spent years being questioned and made to reason aloud has an advantage that no six-week course replicates. It is the same capability the interview tests, by a different route.
The test landscape now
| Test | Run by | Used for | Notes |
|---|---|---|---|
| UCAT | Pearson | Medicine, dentistry across most UK schools | BMAT has been cancelled — UCAT is now the route |
| LNAT | Pearson | Law at Oxford, Cambridge, UCL, LSE and others | Unchanged |
| TMUA | UAT-UK | Mathematics, Computer Science, and maths-adjacent courses including Economics at some universities | Emphasis on mathematical reasoning, not syllabus coverage |
| ESAT | UAT-UK | Engineering and science courses | Modular: Maths 1 compulsory, plus further science/maths modules depending on course |
| TARA | UAT-UK | Economics & Management, PPE, Human Sciences, Psychology, some History joint courses; also adopted by UCL for several programmes | New — subject-neutral reasoning on unfamiliar material |
| STEP | Cambridge Assessment | Cambridge Mathematics, mostly | Usually a condition of an offer, sat in the summer — not a shortlisting test |
Requirements vary by university and course and change frequently. Verify on each course page and the test websites.
Oxford's course-to-test map
Oxford publishes this directly, and it is worth checking rather than assuming:
| Test | Oxford courses |
|---|---|
| TMUA | Mathematics · Mathematics and Statistics · Mathematics and Computer Science · Mathematics and Philosophy · Computer Science · Computer Science and Philosophy |
| ESAT | Engineering Science · Physics · Physics and Philosophy · Biomedical Sciences |
| TARA | Philosophy, Politics and Economics · Economics and Management · Human Sciences · Psychology (Experimental) · Psychology, Philosophy and Linguistics · History and Economics · History and Politics |
| UCAT | Medicine · Graduate-entry Medicine |
| LNAT | Law (Jurisprudence) · Law with Foundation Year |
Courses not listed may require no test. Confirm on the specific course page.
One genuinely useful consequence of the overhaul: because Oxford, Cambridge, Imperial, UCL, LSE and others now use the same UAT-UK tests, a student sits the relevant test once and it serves every application that requires it. Previously an applicant might sit three different exams for three universities.
The offsetting cost, which nobody mentions: the retired Oxford tests had decades of past papers and established preparation routes. The new tests do not. Anyone selling you an extensive back-catalogue of practice material for TARA should be asked where it came from.
The registration trap
This is where Hong Kong families lose places for reasons that have nothing to do with ability.
Nobody registers your child but you. Unlike school exams, admissions tests require the candidate to create an account, register, and book a test slot. Schools frequently do not do this, and do not always remind families.
The pattern of the cycle, using Oxford's published dates for the current October sitting as an illustration:
| Stage | Timing |
|---|---|
| Registration and account creation opens | Early June |
| Test booking opens | Late July |
| Access arrangements deadline | Mid-September |
| Test booking closes | Late September |
| Test sitting | Mid-October |
| Results released to candidates | Mid-November |
Dates are for the current cycle and change annually — confirm at esat-tmua.ac.uk, ucat.ac.uk, and lnat.ac.uk.
The three traps, in order of how often they cost places:
- Booking closes before the UCAS deadline. Test booking typically closes in late September; the Oxbridge and medicine UCAS deadline is 15 October. A family working backwards from the UCAS date has already missed the booking window.
- Test centre availability in Hong Kong. These are delivered at Pearson test centres. Slots in Hong Kong are finite and fill early, and international candidates may face specific date or centre restrictions. Book on the day booking opens, not the week it closes.
- UCAT is taken over the summer, on a different calendar from the UAT-UK tests — typically registered in early summer and sat between July and September, before the school year has properly restarted.
What each test rewards
UCAT — speed under severe time pressure, verbal and quantitative reasoning, decision-making, and situational judgement. The binding constraint is pace, not difficulty. Practice is genuinely useful here because familiarity with format buys time.
LNAT — reading comprehension and argument analysis, plus an essay. Not a law knowledge test; there is nothing to revise. What helps is having read seriously and argued in writing.
TMUA — mathematical thinking and reasoning rather than curriculum coverage. Students who have only ever executed procedures find it disorienting; students used to proof and problem-solving do not.
ESAT — modular science and maths, applied. Content-based, so subject preparation transfers.
TARA — reasoning on unfamiliar material, deliberately subject-neutral. The least "revisable" of all of them, and the most closely aligned with the thinking habits described in our guide on critical thinking.
The Oxbridge interview, honestly
Oxford describes its interviews plainly: an academic conversation about your chosen subject, similar to a short tutorial, designed to explore how you think and engage with new ideas. Interviews are typically conducted by two tutors and, in recent cycles, held online in December, with decisions in January.
That description is not marketing softening. It is a literal account of the format, and it explains almost everything about how to prepare.
What the tutors say they are doing. Oxford's Director of Admissions and Outreach has described the interview as primarily an academic conversation based on a passage of text, a problem set, or technical discussion related to the course — and has stressed that every question has a purpose: to assess how a student thinks about their subject and responds to new information. Tutors have been explicit that the object is not to catch anyone out.
The single most useful piece of advice Oxford gives, and the one Hong Kong students most need: start by making very obvious observations and build from there. Solving the problem quickly matters less than showing how you use information and analysis to get there.
Read that again, because it inverts what most Hong Kong students have been trained to do. A student conditioned to produce the correct answer will sit in silence, working privately toward certainty, and say nothing for ninety seconds. The tutor cannot assess silence. The student who says "well, the obvious thing to notice is…" and then reasons out loud, badly at first, is the one demonstrating exactly what is being measured.
Real questions, and what they are actually testing
These are drawn from questions Oxford has itself published, together with tutors' explanations. Note how ordinary the material is — and how little of it can be revised.
| Subject | The question | What it's really testing |
|---|---|---|
| Biological Sciences | Ladybirds are red. So are strawberries. Why? | Whether you notice that the same trait serves opposite purposes — warning versus attraction. Tutors often start from a specimen or a piece of fruit |
| Biological Sciences | Why does it matter if tigers become extinct? | Whether you can move past "it's sad" to ecosystems, trade-offs, and what conservation is actually for |
| Materials Science | What temperature would a hot air balloon need to be to lift an elephant? | Estimation and structured reasoning with incomplete information — can you build a model rather than wait for data? |
| Experimental Psychology | Why might research suggest older siblings have higher IQs? | Hypothesis generation, and whether you interrogate the finding itself rather than accepting it |
| Medicine | Why is the death rate from cancer higher in the UK than in the Philippines? | Whether you reach for the counter-intuitive explanation — life expectancy, diagnosis rates — rather than assuming worse healthcare |
| English | Why might an English student be interested that a soap opera has run for fifty years? | Whether literary thinking extends to popular culture, form, and audience |
| History | Does the word "political" mean the same thing in different contexts? | Conceptual precision, and comfort with definitional instability |
| Engineering | Explain exactly why a ruler topples over. | Whether you can produce a rigorous account of something entirely familiar |
| Economics | Should governments curtail bankers' bonuses? | Structured argument on a contested question, and whether you can hold both sides |
The pattern across all of them: the material is deliberately ordinary. The assessment is entirely about the reasoning you build on top of it. There is no body of knowledge that prepares you, and that is the point.
On the famous strange questions — the banana, the brick, the various riddles that circulate in parent groups. Some are real, many are embellished, and the mythology does more harm than good, because it frames the interview as a trick to be survived rather than a conversation to be had. Oxford publishes genuine examples with tutor commentary precisely to dispel this. Work from those.
How to prepare — and how preparation backfires
What genuinely helps: - Re-read everything you submitted. Your personal statement and any written work are legitimate territory, and being unable to discuss a book you claimed to have read is the worst available outcome. - Practise thinking aloud. Have someone ask "why do you think that?" and then push back, repeatedly, until reasoning in public stops feeling exposing. - Read beyond the syllabus in your subject, and form views about what you read. - Work through the published sample questions — not to memorise answers, but to notice how the reasoning develops. - Get comfortable being wrong out loud. Tutors routinely push a candidate to see whether they revise their thinking or defend it rigidly. Both can be right; refusing to engage is not.
What backfires, and is visible immediately: - Rehearsed answers. Tutors interview hundreds of candidates. A memorised response is obvious, and the follow-up question exposes it instantly. - Performing confidence. Fluency without substance is transparent and worse than honest uncertainty. - Refusing to guess. Silence is the least informative thing a candidate can offer. - Over-coaching. A heavily drilled candidate frequently underperforms an unprepared one, because the drilling replaces thinking with retrieval.
Where Hong Kong families most often go wrong
- Preparing for retired tests. MAT, PAT and TSA tutoring is still being sold. Check what your child's course actually requires, this cycle.
- Missing the booking window. Test booking closes weeks before the UCAS deadline. This is the single most common avoidable failure.
- Assuming the school handles registration. It usually doesn't.
- Treating tests as revision exercises. For TARA and TMUA especially, there is little to revise — the capability is built over years.
- Drilling the interview. The most over-prepared candidates often present worst.
- Discovering the requirement in September. Whether a course needs a test should be known when the list is built, not after.
Your test and interview planner
| Course & university | Test required | Registration opens | Booking closes | Test date | Interview? | Date checked |
|---|---|---|---|---|---|---|
- [ ] Have we checked each course's requirement on the university's own page?
- [ ] Have we registered on the test provider's site — not assumed the school would?
- [ ] Have we booked a Hong Kong test centre slot early?
- [ ] Do we know whether the course also requires written work?
- [ ] Have we diarised the booking close date, not just the test date?
Before the autumn: a final check
- [ ] Do we know which test, for which course, in this cycle?
- [ ] Is the preparation aimed at a test that still exists?
- [ ] Is the booking made, with a confirmed centre and date?
- [ ] Can our child reason aloud under questioning — not just solve problems privately?
- [ ] Have they re-read their personal statement and written work?
- [ ] Do they understand that starting with the obvious is the right opening move?
What to do next
- Verify the test requirement today, at source. Given how much changed for 2026, assume any advice older than one cycle is wrong until confirmed.
- Diarise the booking-close date, not the test date. That single calendar entry prevents the most common failure in this process.
- Practise thinking out loud, starting now. The tests and the interview measure the same underlying capability, and it is built over years — see our guide on critical thinking and communication.
Admissions tests and interviews reward a way of thinking that is built long before the autumn of the application year — and the requirements changed substantially for 2026. If you'd like an honest read on what your child's courses require and how ready they are, our founders offer a free Preliminary Fit Review — a short conversation about fit and next steps, with no obligation.
This guide is for general information and describes broad patterns, not the requirements of any specific university or course. Admissions tests, formats, registration windows, and interview practices change frequently and vary by course; always verify current requirements directly with UAT-UK, UCAT, LNAT, and each university before applying. Sample interview questions are drawn from material published by the University of Oxford and are illustrative of format only. References describe the admissions landscape and are not recommendations, rankings, or guarantees of any outcome.
Want an honest read on your own child?
This guide is general information. For a personal view on fit and next steps, take the free Fit Check or request a confidential Preliminary Fit Review.