Oxford vs Cambridge Interviews: The Key Differences by Subject for 2027 Entry

Oxford will interview every shortlisted candidate online in December 2026 and announce decisions on 12 January 2027. Cambridge lets each College decide the format, runs its main window from 7 to 18 December 2026, and then pools. This page compares what the two universities have actually published for 2027 entry — shortlisting rates, test changes, written work and the winter pool — rather than what the interview room supposedly feels like. Written practice across 17 subjects is in the expert interview packs, £180 each.

What have Oxford and Cambridge actually published for 2027 entry?

Most comparisons of these two interviews are written in adjectives. Oxford is tutorial-style, Cambridge is collaborative, one is combative and the other is warm. None of that is checkable, and none of it helps a candidate decide what to do in the second week of December. Everything below is built from what the two universities have themselves published for the 2027-entry cycle, and the pattern that emerges is not the one the adjectives predict: the sharpest differences have moved out of the interview room and into the paperwork around it — who chooses the format, which test you sit and when, what you send in advance, and what happens in January.

Format first. Oxford's guide for applicants is unambiguous: "Shortlisted applicants will be invited to online interviews in December 2026. Oxford interviews are academic conversations about your chosen subject, similar to a short tutorial and designed to explore how you think and engage with new ideas." There is no in-person alternative to weigh up. Cambridge has made no such single decision. Its interview page states that interviews "may be online or in person, depending on which College is assessing your application and the course you have applied for". One university decided once, for everybody; the other left the decision with its Colleges.

The calendars differ too, and not in the way most guides describe them. Cambridge publishes a defined main window — 7 December to 18 December 2026 — with a winter pool in the new year. Oxford publishes a December 2026 period and one hard date at the end of it: "Shortlisted candidates for entry in 2027 will be informed of the outcome of their application on 12 January 2027 via UCAS. Colleges will follow up directly later that day." What neither university publishes is the "two to three days" of interviews that circulates online, and neither publishes a length for an individual interview. Cambridge publishes a total only: "Most applicants will have 1 or 2 interviews lasting a total of 35 minutes to an hour."

Volume is the fourth. Oxford tells shortlisted candidates to "remember that Oxford typically offers around 20,000 online interviews to approximately 10,000 shortlisted candidates" — 20,000 interviews spread across 10,000 people is two interviews each on average, university-wide. And Cambridge's panels are slightly larger than Oxford's rather than smaller: Cambridge describes candidates discussing "topics with two or three interviewers", while Oxford states that "typically, interviews are conducted by two tutors, although there may occasionally be more. If you are applying for a joint course, you should expect to be interviewed by tutors representing each subject."

Read down that table and the honest summary is that the interview itself is the part the two universities describe most similarly: a subject conversation, with two or three academics, lasting well under an hour in total. It is everything bracketing the interview where they genuinely diverge, and almost all of it is decided before December.

Online or in person: which one applies to you?

For a family in the UK this is not a question about atmosphere. It decides whether somebody takes days off work in the second week of December, whether there is a train and a hotel to book, and whether a candidate sits the most consequential conversation of their year at the kitchen table or in a room they have never seen. It deserves a straight answer well before December, and at one of the two universities there is one.

At Oxford the answer is the same for every candidate and every subject: online, in December 2026, conducted by two tutors, with tutors representing each subject if the application is for a joint course. Nothing about the College named on the UCAS form changes it, and nothing about the subject changes it either. There is no travel decision to make.

At Cambridge the answer is set College by College. We read the University's published interview-format table on 23 August 2026 and worked through every College listed on it. The largest group commit to online interviews with no condition attached. A much smaller group commit to in-person interviews with no condition attached. The remainder attach a condition, and for most of those the condition is where the applicant is based rather than which subject they have applied for.

Two cautions about that grouping. It is Cambridge's table, not ours — we have only sorted what was published on the date shown, and a College can revise its entry before the cycle opens. And it is the College assessing your application that matters, which is not always the College you named on UCAS. If your application moves in the pool, the format can move with it. Check the entry for your own College before you book anything, and treat the second week of December as unavailable until you have.

One correction is worth making explicitly here, because the opposite is widely published: it is Oxford that specifies two interviewers and Cambridge that specifies two or three. Guides describing a lone Cambridge don facing a candidate across a desk are describing something Cambridge does not publish. If a mock interview is being run for a Cambridge candidate, it should be run with a panel, and for an Oxford candidate it should be run on screen, because a camera changes how a conversation feels far more than the wording of the questions does.

How many applicants does each university actually interview?

Oxford publishes a figure on every course page that Cambridge does not publish at all. Each Oxford course shows an interviewed percentage, a success percentage and an intake, and each is labelled as a three-year average covering 2023–25. It is the most useful number on this page, because it tells a candidate how much of the decision has already been taken before anybody speaks to them.

University-wide, Oxford states that it "typically receives over 23,000 applications for around 3,300 places every year and shortlists approximately 10,000 candidates". The most recent published outcome is close to that planning figure: "In 2025, 3,302 students were admitted to Oxford to begin their undergraduate studies; 79% of undergraduate students admitted were from the UK." Those aggregates, though, conceal a spread wide enough to change what preparation ought to look like.

History interviews 74% of its applicants and English 67%. Medicine interviews 29%. The success rates do not spread anything like as far — 23% for History, 22% for English, 11% for Medicine — so the difference is not that History is easier to get into. It is that History and English do almost none of their filtering before the interview, and Medicine does most of it. Nothing in that data says one subject is gentler than another; it says they put the same squeeze in different places.

The consequence for preparation is direct. A History or English candidate should assume from the day they apply that an interview will happen, because on these figures most applicants get one, and the interview therefore carries close to the whole decision. A Medicine, Maths or Physics candidate is preparing for two events with entirely different failure modes: a test in October that decides whether there is an interview at all, and then a conversation with a field already thinned to roughly a third of the applicants. Spending the autumn on interview technique when the October test is the real gate is the commonest planning error we see.

Cambridge publishes no equivalent per-course rate. What it publishes is a competition ratio: "On average across all subjects, we receive about six applications per place." That answers a different question and cannot be set beside Oxford's interviewed percentage. If you have seen a table claiming Cambridge interviews a given percentage of applicants in a named subject, it did not come from Cambridge. Working out which of these two shapes a candidate is actually facing is the first step, because the published data supports a different plan for History than for Medicine.

What does the Cambridge winter pool actually do in January?

The winter pool is the part of the Cambridge process most often described as Oxford's system with a different name. Cambridge's own decisions page shows it is not. "Around 18% of applications received in October 2025 were placed in the winter pool." Of those, "In January 2026, 1060 pooled applicants (around 1 in 4 pooled applicants) were offered a place by a different College to the one they applied to." And then the figure that reframes everything: "In January 2026, 20 applicants were invited to a second interview."

Set the last two figures against each other. 1060 offers made from a different College, 20 second interviews issued — 1060 divided by 20 is 53, so for every pooled applicant called back to Cambridge in January, fifty-three received an offer from another College without being interviewed a second time. Being pooled at Cambridge is, in practice, a paper process. Your file moves; you almost certainly do not. That should change how a pooled candidate spends January: there is far more value in getting on with A-level study than in drilling for an interview that, on Cambridge's own numbers, is very unlikely to be offered.

Oxford's equivalent runs inside the December week instead, and it can arrive with very little warning: "In other subjects, you might have 'initial' interviews at your first college and then be invited to an 'additional' interview at another college after these have taken place. We aim to give everyone a minimum of 24 hours' notice of these additional interviews." Oxford calls this reallocation, and the additional interview is a normal part of the December fortnight rather than an exception. The term "winter pool" belongs to Cambridge, and applying it to Oxford implies a January round that Oxford does not run.

The practical difference is a diary one, and it is the sort of thing families discover too late. A Cambridge candidate whose last interview falls on 18 December is, for almost all purposes, finished until the outcome arrives. An Oxford candidate is not finished until the December window closes, because a second College can appear with a day's notice and expect a full interview at short order. If there is travel planned in that fortnight, plan it around Oxford's whole window rather than around the date printed on the first invitation.

Which admissions tests survive for 2027 entry?

This is where the two universities have moved furthest, and where most published comparisons are now simply out of date. Oxford has handed its testing to UAT-UK. Its admissions-tests page states that "the ESAT (Engineering and Science Admissions Test), TARA (Test of Academic Reasoning for Admissions) and TMUA (Test of Mathematics for University Admission) will be taken by Oxford candidates where required for admission to their chosen course", alongside the LNAT for Law and the UCAT for Medicine. The papers Oxford wrote and ran itself do not appear on that list at all.

The consequence for two subjects is dramatic and almost nobody has written it down. Oxford English now says, in terms, "You do not need to take a written test as part of an application for this course." A subject that for years was gated by a written paper is now gated by an essay the candidate has already written and a conversation in December. Oxford History is in the same position — nothing resembling its old paper appears among the tests Oxford lists for 2027 entry. Physics candidates now sit the ESAT where they once sat a Physics-specific paper, and PPE candidates sit TARA.

The dates are published, and they are earlier than families expect. "Test booking closes 28 September 2026 (6pm UK time) Test dates 12-16 October 2026 Candidates receive test results via their UAT-UK account 16 November 2026". Booking therefore closes before many schools have finished their UCAS references, and an unbooked test is an ended application. Of everything on this page, that deadline is the one with the most catastrophic and most avoidable failure attached to it.

Cambridge uses the same UAT-UK tests where they apply, and publishes their shape. The ESAT "is a computer-based assessment. It is made up of individual multiple-choice modules lasting 40 minutes each. You will sit these back-to-back on the day of the test." The TMUA "lasts 2 hours 30 minutes in total. It is divided into two parts: Applications of Mathematical Knowledge and Mathematical Reasoning." Candidates in some regions have less choice of day than the published window suggests: for the ESAT, "Applicants from China, Hong Kong and Macau must take the test on either the 12 or 13 October", and for the TMUA, "Applicants from China, Hong Kong and Macau must take the test on either the 15 or 16 October". Families applying from Hong Kong or the mainland should fix those dates before anything else in the autumn calendar. Our admissions test preparation is built around that October window rather than around the interview, because for several of these subjects the test is the harder gate.

Cambridge College assessments: the test you sit after you are shortlisted

Where Oxford moved its testing earlier and handed it to a third party, Cambridge moved part of its testing later and kept it in-house. College admission assessments carry no registration deadline anybody can miss: "If you're invited for interview, the College that is interviewing you will arrange for you to take the relevant College admission assessment. They will let you know when and how you should take your assessment." There is nothing to book. There is also, in a good many subjects, nothing to sit — Cambridge English states plainly that "there is no admissions assessment for this course for the majority of Cambridge Colleges".

Where they do exist they are specified in unusual detail, which makes them straightforward to prepare for. The Linguistics assessment "consists of three parts, which should each take about 20 minutes. You will have 60 minutes to complete the assessment. Each part is worth 30 marks, giving a 90-mark maximum for the overall written test." Three parts at 30 marks each is exactly the 90-mark maximum Cambridge quotes, and the timing deserves reading twice: three parts of about 20 minutes each fills 60 minutes of a 60-minute paper, so there is no slack for a slow start. The Modern and Medieval Languages assessment opens with a "discursive response in Foreign Language (40 minutes)".

Computer Science carries the clearest example of College-level variation anywhere in the Cambridge system: "If you're applying to Peterhouse or Trinity you will also need to take the Computer Sciences Aptitude Test (CSAT)", on top of the TMUA that every Cambridge Computer Science applicant sits. Two named Colleges, one extra paper. That is a College choice with a real workload attached, and it is invisible unless somebody reads the admission-tests page as well as the course page.

The structural point is worth stating plainly, because not one competing page on this search result carries it. At Oxford, everything testable happens before shortlisting, in a fixed October window, run by an external body. At Cambridge, the pre-registration tests are the same four — ESAT, TMUA, UCAT and LNAT — but a second layer sits after shortlisting, arranged by whichever College is assessing the application, with no registration and, in several subjects, no assessment at all. Any preparation calendar that treats "Oxbridge admissions tests" as a single thing gets the timing wrong in both directions: too late for Oxford, and too anxious about a Cambridge paper that may never be set.

Written work: one capped essay at Oxford, two pieces at Cambridge

No competing page on this search result covers written work at all, which is strange, because it is the cleanest contrast available and it carries the earliest real deadline. Oxford centralises it. For History, "the essay should be a maximum of 2000 words. Extracts from longer essays or pieces of project work are permissible, but should be selected so that they can stand alone and show the applicant's ability to make an argument", and the University sets a single deadline for it: 10 November 2026. One piece, one word limit, one date, the same for everyone.

Cambridge devolves it, and asks for more. For History, "you will need to submit 2 pieces of written work at all Colleges except Sidney Sussex. If you are applying to Sidney Sussex, you will need to submit one piece of written work." Cambridge English also asks for two pieces. The deadline and the format come from the College rather than the University, which means there is no single date to write in the diary — there is your College's date, and finding it is the applicant's job.

The choice of essay matters more than most candidates assume, because both universities treat submitted work as material that can be discussed in the room. An essay you cannot defend under questioning is worse than a slightly less polished essay you can. In practice the strongest submissions come from ordinary A-level coursework the candidate genuinely argued with, not from something written specially in October to impress an admissions tutor. Oxford's wording about extracts standing alone is a heavy hint about what is being assessed: whether you can make an argument, not whether you can cover a topic.

Two subjects break the pattern in opposite directions. Cambridge Medicine asks for no written work before interview at all. Oxford Music asks for something no other subject asks for, and only once a candidate has been shortlisted: "If you are shortlisted for interview for Music you will be asked to submit a 5-minute (max) continuous video recording of a musical performance on your chosen instrument or voice." Continuous is the operative word. It is a performance under pressure, not an edit, and it needs rehearsing as a single take.

Medicine: where the two systems diverge most

Medicine is the subject where the published differences are widest, and it is also where the most misinformation circulates — including, until this rewrite, on this page. Oxford uses the UCAT for Medicine. It does not use any Oxford-specific medical admissions test, and no such test exists for 2027 entry. Anything you read naming an Oxford-only medicine paper is describing something that has never been published by the university.

The Oxford shortlist is small in absolute terms as well as in percentage terms: "Please note that competition to study Medicine at Oxford is particularly strong and only around 425 applicants are shortlisted for interview each year." That maps onto the published three-year average for 2023–25 of a 29% interviewed rate, an 11% success rate and an intake of 155. Every shortlisted candidate then sees two Colleges. On the timetable Oxford published for the previous cycle — the 2027-entry subject timetables were not yet available when we checked — "each shortlisted candidate will be interviewed by two colleges, with the interviews being either on Monday 15 or on Tuesday 16 December." Treat the specific day as last cycle's pattern rather than a promise for 2027, but treat the structure as the shape to prepare for: two Colleges, compressed into a single day, with no additional round afterwards.

Cambridge starts from the same test and reads it differently. "For UCAT results, we will look at an applicant's overall cognitive subtest score. We will not be using the situational judgement score as part of our assessment for 2027 entry." For anyone who has been working through situational judgement material on the assumption that all of the UCAT counts everywhere, that is a genuine change of plan: at Cambridge, for this cycle, it does not count. Cambridge also asks Medicine applicants for no written work before interview, which removes an autumn task that History and English applicants cannot avoid.

So the two Medicine routes look like this. At Oxford: a UCAT, a shortlist of around 425 people, and two College interviews compressed into one December day, all of it online. At Cambridge: a UCAT read for its cognitive score only, no written work, and an interview format decided by whichever College is assessing the application. A candidate cannot hedge across both, and the preparation is not interchangeable. Where the real gap is scientific depth rather than interview technique, our subject tutors work on the science first, because no amount of interview coaching rescues a candidate who cannot reason through unfamiliar material out loud.

You can only apply to one, so prepare for the institution rather than for "Oxbridge"

Oxford states the rule on its own UCAS guidance page: "You can only apply to one course at Oxford, and you can't apply to Oxford and Cambridge in the same year." That one sentence is why this comparison matters at all. There is no hedging strategy, and no shared preparation plan that genuinely covers both, because from the moment the UCAS form is submitted only one of these two processes is ever going to happen.

Look at what that single choice determines. Whether anybody travels in December. Whether the test falls in October, or after shortlisting, or does not exist. Whether the candidate submits one essay capped at 2000 words by 10 November 2026, or two pieces on a College's own timetable. Whether an unexpected second interview is a December possibility with a day's notice or a January near-impossibility. Whether the answer arrives on 12 January 2027 or waits on a pool decision. Not one of those is a matter of interview style, and every one of them is published.

Which is not to say style is irrelevant — it is to say that nothing either university publishes supports the confident subject-by-subject characterisations that fill this search result: that Oxford Maths works from first principles while Cambridge scaffolds, that Oxford PPE is combative while Cambridge HSPS is collaborative. Our tutors who have sat on both sides of these conversations do observe patterns, and we will tell candidates what they observe, labelled as observation. What we will not do is print it as though a university had published it. Neither university publishes a weighting for the interview against the rest of the application either, so anyone quoting a specific proportion is guessing.

One further change to the preparation landscape is not widely reflected yet. Cambridge now states that "Sharing admissions assessment or interview questions is strictly prohibited. You should not share this information during or after the admissions cycle. You may disadvantage your application by sharing assessment or interview questions." Question banks mined from student forums were always unreliable; using them now runs against a published rule, and the rule names a consequence. The official sources are better and far more current anyway: Cambridge's per-College format table, Oxford's subject timetables, and the course pages both universities revise each summer. That is where every figure on this page came from, and it is where the 2027 versions will appear first.

Frequently Asked Questions