Year 13: A-Level, IB and University Preparation for 2027 Entry

Most of this page applies whether or not Oxbridge is part of your plan. One section does not: if Oxford or Cambridge interviews are on your calendar, our Oxbridge interview question packs let you work through real subject questions before the invitation arrives, at £180 each – useful for that minority of Year 13 readers, and not something the rest of this page assumes you need.

Most Year 13 guidance is undated. This is not. It carries the 2027-cycle calendar as UCAS publishes it – including the 13 January 2027 equal consideration date that replaced the 31 January everyone still quotes – the corrected Oxford and Cambridge admissions test maps, the booking deadlines that closed in September, and published competition data: applicants per place, offer conversion, and what admitted students actually achieved.

What did summer 2026 actually show?

The top of the A level distribution barely moved this summer. JCQ recorded 28.5% of A level results awarded A*–A, compared to 28.3% in 2025, and 9.6% awarded A*, compared to 9.4%. The overall pass rate held exactly: 97.5% of results were awarded A*–E, the same as in 2025. Subtract one from the other and the movement at the top is 0.2 percentage points on A*–A and 0.2 on A* – noise, not a trend. AS levels drifted the other way by exactly that margin, from 24.9% awarded A down to 24.7%. Nothing in that data says A levels got harder or easier. It says the standard sat still, which is the most useful thing a Year 13 student can know when setting a target grade in September.

The queue behind that standard did not sit still. 349,917 students took A levels in 2026 against 341,577 in 2025, a rise of 2.4%, and entries at all ages rose 2.7%, from 882,653 to 906,425. Divide one by the other – 906,425 entries across 349,917 students – and the average student sat 2.6 A levels. Set against that, the population of 18 year-olds increased by 3.9%. Entries grew 2.7% while the cohort that generates them grew 3.9%, a gap of 1.2 percentage points. More young people are competing for university places than the entry figures on their own suggest.

That is what should change how a Year 13 student reads a conditional offer. 9.6% of 906,425 entries works out at roughly 87,000 grades at A*, which sounds like a great many until you set it against the number of places on the courses that ask for them, further down this page. Grades did not inflate and they did not deflate; the competition around them thickened slightly. If your year is going to be built around A-level tuition aimed squarely at the grades written into your offer, that arithmetic is the argument for starting in the autumn rather than after the January deadline has passed.

Which dates in the 2027 cycle actually bind?

The most repeated wrong fact in Year 13 guidance – including, until this revision, on this page – is that the main UCAS deadline is 31 January. It is not. For 2027 entry the equal consideration date for most undergraduate courses is 13 January 2027 at 18:00 UK time. UCAS publishes it plainly: "2027 Jan 13 (18:00 UK time) Equal consideration date for applications for most undergraduate courses". If a school handbook, a printed planner or a blog post still says 31 January, it is describing a cycle that no longer exists, and it is handing students an extra fortnight they do not have.

Two dates come before it and get missed because they are not deadlines in the ordinary sense. Applications for 2027 entry opened on 12 May 2026, but completed applications could not be submitted to UCAS until 1 September 2026. The window between those two dates is registration and drafting time rather than dead time, and the students who used it are the ones who were not writing three personal statement answers in the same fortnight they sat an admissions test. The 15 October 2026 gate, also at 18:00 UK time, covers Oxford, Cambridge and most courses in medicine, dentistry and veterinary medicine or science. It is the same date for all of them, and it arrives before most of Year 13 has been taught.

The two late dates are worth knowing even if you never use them. The final date for any 2027 entry application is 23 September 2027 at 18:00 UK time, and the last date to add a Clearing choice is 18 October 2027. Neither is a plan. Both are the reason it is worth deciding, calmly and early, what your insurance choice would actually be, because the alternative is deciding it on results day in a phone queue. If you are still weighing sixth form routes rather than university ones, our sixth form entry guidance covers the stage immediately before this one, and the two decisions are more connected than they look.

Why is the personal statement three questions now, not an essay?

UCAS changed the shape of the personal statement from 2026 entry onwards, and a great deal of advice still circulating describes the old one. It is no longer a single free-standing essay. It is three prescribed questions, answered separately in the application. Question 1 is "Why do you want to study this course or subject?". Question 2 asks how your qualifications and studies have prepared you for it. Question 3 asks what else you have done outside education to prepare, and why that experience is useful. The reader is no longer hunting through prose for those three things; they are asking for them directly.

The limits are where people get caught out. Each answer has a minimum character count of 350 characters, labelled on the question box, and the overall limit is unchanged at 4,000 characters including spaces. Three minimums of 350 characters account for 1,050 characters, which leaves 2,950 characters to distribute however your case demands. That is a genuine editorial decision rather than a formality. A candidate for a course with a heavy subject test can reasonably spend more of the 4,000 on question 2, showing how the syllabus feeds the test. A candidate whose case rests on serious independent work outside school needs room in question 3.

The practical consequence is that the old advice to open with a hook is close to useless now. Question 1 already tells the reader what the paragraph is about, so an arresting first line about a childhood fascination is spending characters on something the form has already established. What survives the change is specificity: the named book, the named module, the thing you actually did, and the conclusion you drew from it. Our subject specialists are usually more useful on question 2 than a general application coach, because the strongest answers there read like a student describing their own syllabus honestly rather than praising it.

Which admissions test does your course actually need?

This is where the previous version of this page was wrong twice, and where most competing guides are simply silent. For 2027 entry Oxford uses the UAT-UK suite – the ESAT (Engineering and Science Admissions Test), TARA (Test of Academic Reasoning for Admissions) and TMUA (Test of Mathematics for University Admission) – alongside the LNAT and the UCAT. TMUA stands for Test of Mathematics for University Admission and nothing else. The MAT does not appear in Oxford's list for 2027 entry, and Oxford Mathematics now sits the TMUA. TARA is the test almost nobody writes about, and it covers a very large and very popular group of courses.

Cambridge's map is different, and this is the correction that matters most, because getting it wrong means preparing for the wrong test. Cambridge Engineering does not use the TMUA. It uses the ESAT, and so do Natural Sciences, Chemical Engineering and Biotechnology, and Veterinary Medicine. Cambridge uses the TMUA for Mathematics, Computer Science and Economics. Computer Science applicants to two colleges carry a second hurdle: Cambridge states that if you are applying to Peterhouse or Trinity "you will also need to take the Computer Sciences Aptitude Test (CSAT)". That is a college-level requirement layered on top of a university-level one, and it is easy to read past.

One piece of genuine good news sits inside the UAT-UK arrangement. Oxford states that applicants who also apply to other universities using these tests "will only need to take the relevant UAT-UK test once". A student applying to Oxford and Cambridge for a physical sciences course sits one ESAT, not one per university. Cambridge also divides its tests into two categories that need different planning: pre-shortlisting tests, which you register for in advance and take at a test centre near you, and college assessments arranged after an interview invitation, which need no registration at all. Preparing for the two is not the same work, and our admissions test preparation treats them as separate jobs.

When do the test windows actually close?

The trap of Year 13 is not the deadline everyone remembers. It is the booking deadline weeks earlier that nobody writes in a planner. In this cycle the UCAT booking deadline was 16 September 2026 at 15:00 UK time, registration closed at the same moment, and the last test day was 24 September 2026. The UCAS gate those scores feed into is 15 October 2026. That is 21 days between the last chance to sit the test and the deadline the test exists to serve, and a student who decides in early October to apply for medicine has already lost the year.

The UAT-UK October sitting ran on its own clock, published by Oxford. Booking opened on 20 July 2026 at 3pm UK time. The deadline for requesting access arrangements was 14 September 2026 at 6pm, the deadline for requesting a bursary was 21 September 2026 at 6pm, and booking closed on 28 September 2026 at 6pm. The tests themselves ran from 12 to 16 October 2026, and candidates received their results through their UAT-UK account on 16 November 2026. Read that sequence backwards and the shape of the autumn becomes obvious: the decision to apply to Oxford has to be made in the summer, because the booking closes before the teaching term has properly started.

The UCAT itself is a large, blunt instrument. Around 37,000 candidates sit it each year in Pearson test centres throughout the UK and worldwide. It produces a total score of 900–2700 plus a Situational Judgement Band, and it is marked on the number of correct answers with no negative marking for incorrect ones. That last detail changes tactics more than most candidates realise: there is never a reason to leave an answer blank, and time spent agonising over a question you cannot do is time taken from questions you can. Booking is done by the student directly through the official UCAT website, not by a school.

Why does the LNAT have more than one deadline?

A law applicant with five choices can be running three different clocks at once, and nothing in the process sends a reminder that they are different. Registration and booking for entry in September 2027 opened on 1 August 2026, and testing started on 1 September 2026. From there, the date by which you must have sat the test depends entirely on where you have applied – not on when your UCAS application went in.

The one that catches people is 31 December 2026. The LNAT states it flatly – "31 December 2026: Sit the LNAT before or on 31 December 2026" – and that date falls inside the Christmas holidays, in the middle of mock season, at the point when a student who is thinking in January-deadline mode has not looked at the test for weeks. The Oxford and Cambridge clock is far earlier and aligns with the 15 October 2026 UCAS gate, which at least has the merit of being the date they were already working towards.

One further rule deserves a line of its own: an LNAT result cannot be carried forward from one academic year to the next. A candidate who sits it, does not get the offer they wanted, and reapplies the following year sits the entire test again from scratch. That is worth knowing before anybody decides to treat a first sitting as a low-stakes rehearsal, because there is no such thing as a low-stakes sitting here. It also means preparation is not an investment you bank; it is work that has to be current in the cycle you are applying in.

What did admitted students actually get?

Competitive is an adjective. Published attainment is a number, and the distance between the two is the most useful thing on this page. Oxford's own undergraduate admissions report gives the 2025 cycle as 23,329 applications, 3,891 offers and 3,302 students admitted. That is 23,329 applications against 3,891 offers, or about one offer for every six applications, and 3,302 of those 3,891 offers turned into places, a conversion of 84.9%. Almost everyone who gets an offer takes it and meets it, which means the real contest happens before the offer, not after it.

Then the grades. Among UK students admitted in 2025, 87.2% achieved A*AA or better at A-level and 47.1% achieved three A* grades or better. Apply the second figure to the 3,302 admitted and roughly 1,555 students arrived with at least three A*s. Oxford's published minimum offers sit below that for most courses. The distribution of people who actually got in does not. This is not an argument that the published offer is dishonest; it is an argument that meeting the published offer and being competitive for it are two different exercises, and only one of them is written on the course page.

Oxford Mathematics illustrates it cleanly. The published offer is A*A*A with the A*s in Mathematics and Further Mathematics; where Further Mathematics is not available it is either A*AAa with an a in AS-level Further Mathematics, or A*AA with the A* in Mathematics. On Oxford's published three-year figures, 32% of applicants were interviewed and 9% were successful, against an intake of 183. Divide 32 by 9 and roughly one in every 3.6 interviewed candidates receives an offer. The interview is not a formality that follows the shortlist; it is a second, comparably severe filter.

Read the right-hand column against the second one. Cambridge asks for A*A*A in Mathematics, and 88% of entrants held at least A*A*A*. It asks for A*A*A in Medicine, and 82% of entrants held at least A*A*A*. Cambridge publishes those figures itself, which is more candour than most institutions offer, and a Year 13 student should treat them as the actual target rather than as trivia. The published offer tells you the floor of what is acceptable. The entrant profile tells you what the people sitting next to you in the interview will be predicted.

How oversubscribed is the course you want?

Five UCAS choices are a portfolio decision, and Oxford is one of the few universities that publishes the number you need in order to make it properly: applicants per place, course by course. Most students choose between courses on interest and predicted grades alone, then discover the competition ratio afterwards, if at all.

The spread is wider than students assume. Economics and Management runs at 18.8 applicants per place and Experimental Psychology at 8.6; 18.8 divided by 8.6 is 2.2, so one is more than twice as contested as the other, and a great many applicants treat them as equivalently hard. Computer Science at 14.8 is more contested than Medicine at 9.4, which surprises almost everybody who has absorbed the cultural idea that medicine is the hardest thing to get into. Mathematics at 10.4 and Mathematics and Computer Science at 10.3 are, on this measure, the same decision. None of this should decide what you study. It should decide how you spread five choices.

The funnel behind those ratios is public too. Oxford states that it "typically receives over 23,000 applications for around 3,300 places every year and shortlists approximately 10,000 candidates". Over 23,000 applications for around 3,300 places is roughly one place for every seven applicants, and well over half of all applicants are cut before anyone interviews them. That cut is made on the written application, the predicted grades and the test score – which is how a test whose booking closed in September quietly decides an outcome that is announced in January.

What is STEP, and why does it arrive after the offer?

The previous version of this page discussed Cambridge mathematics admissions without mentioning STEP once. That is a serious omission, because STEP is the condition that actually determines whether a Cambridge mathematics offer becomes a place. Cambridge states that "all Colleges usually require at least grade 1 in two STEP papers (STEP 2 and 3)". A student can hit every A-level condition in the offer and still not take up the place if the STEP grades do not come.

The timing is the part that reshapes an entire Year 13 plan. Cambridge is explicit: "You don't need to take STEP before you apply. You will be asked to sit STEP as part of your offer conditions." So STEP is not an autumn entrance test sat next to the TMUA. It lands in the summer exam season, in the same weeks as A-level Mathematics and Further Mathematics papers, for a candidate who has just spent December on interviews and January on mocks. Cambridge also operates a flexible offer scheme that trades A-level grades against STEP grades; the terms are published on the course page and are worth reading before you assume the standard offer is the one that applies to you.

In practice that means STEP preparation cannot sensibly start in April. The papers reward a way of working that is different in kind from A-level questions – long, open, multi-step problems with no scaffolding, where the first twenty minutes are spent deciding what the question is actually asking. That transition takes months rather than weeks, and it is the reason a serious Cambridge mathematics candidate should treat STEP as a third subject running from the autumn. It is also, usefully, the best possible preparation for the mathematics they will meet in the first year, which is not true of most exam preparation.

What actually happens at interview, and when?

Oxford's interviews for 2027 entry are online. The university states that "shortlisted applicants will be invited to online interviews in December 2026". That is worth saying plainly, because families still plan travel, accommodation and time off work around a December trip to a college that no longer needs to happen – and because an online interview is a different performance from a room-based one, with its own rehearsal requirements around cameras, whiteboards and shared documents.

The notice period is short. Oxford warns that "you may only be given a week's notice that you have been shortlisted", with invitations issued from the middle of November into early December. A candidate who has done no interview preparation before the invitation arrives has, in the worst case, seven days – and those seven days sit inside the mock season. The only workable answer is to prepare on the assumption of being shortlisted, from November, and to accept that the preparation is worth doing anyway because it is mostly subject reading.

The volume tells you what the interview is for. Oxford offers "around 20,000 online interviews to approximately 10,000 shortlisted candidates", which is 20,000 across 10,000 candidates, or two interviews each on average. Nobody is being screened out on the strength of a single conversation, and nobody is being admitted on one either. The interview is a repeated sample of how a candidate thinks when they do not already know the answer. That is why rehearsed answers perform badly, why being wrong out loud and then correcting yourself performs well, and why the most valuable preparation is reading a subject beyond the syllabus rather than drilling questions.

What does an IB score in the forties actually mean?

The IB half of most Year 13 guidance is assertion. A points range for Oxbridge and another for the Russell Group appear everywhere and are sourced nowhere; this page carried a pair of them until this revision. The IB's own provisional statistical bulletin for the May 2026 session replaces that hand-waving with a measurable share of a real, global cohort, which is a far better way to understand what an offer is asking for.

Put the top two bands together. 21.2% of candidates scored 35–39 points and 10.1% scored 40–45, so 31.3% of the May 2026 cohort finished at 35 points or above. Applied to the entry, 10.1% of 209,607 students is about 21,170 candidates worldwide in the 40–45 band. That is the entire global pool from which every university making an offer at that level is recruiting in the same year, across every country and every subject. It makes an offer in the low forties legible: it is asking for roughly the top tenth of the world's Diploma candidates.

The trend underneath the headline is more interesting than the headline. Five sessions ago the 40–45 band held 14.7% of candidates. It then fell to 8.9%, and has climbed back through 9.3% and 9.8% to 10.1%. The pass rate traced the same shape: 86.1% five sessions ago, a low of 80.0%, then 81.9% and now 82.6%. Nothing in that says the Diploma got harder. It says a disrupted assessment cycle inflated the top band, that inflation has been unwound, and the top band is still meaningfully smaller than it was when the current crop of published offer requirements was written.

The core is where points are quietly won and lost. The Extended Essay and Theory of Knowledge are worth up to three bonus points between them, and the share of candidates taking nothing at all from the core has fallen from a peak of 22.1% to 14.8%, a drop of 7.3 percentage points. A student sitting two points short of an offer who treated the Extended Essay as an administrative chore has left the cheapest available points on the table, and they are the only points in the Diploma that can be earned a full year before the exams.

What changed for medicine applicants this cycle?

One scoring rule moved, and it changes what a medicine candidate should be optimising. Cambridge states: "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." Cambridge applies no minimum UCAT threshold either, assessing the academic record, the UCAT and the interview together rather than filtering on a cut-off. Those two facts pull in the same direction: at Cambridge, the cognitive subtests carry the weight.

That is a narrower instruction than it first sounds, and it should not be over-read. It does not mean the situational judgement section is optional – you still sit it, and other medical schools still read it, some of them heavily. It means that for one institution, in one cycle, hours spent grinding situational judgement practice buy nothing there. A candidate applying to Cambridge and to three other medical schools has to work out whose rules bind them where, and to allocate preparation accordingly instead of treating the UCAT as a single undifferentiated score.

Two other points belong here. Imperial College London is a UCAT Consortium university for 2027 entry, so the UCAT is the test that applies there now. You will read elsewhere that this followed the abolition of the BMAT in 2023; we could not source that date from a primary publication, so this page states only the part that is verifiable and leaves the rest alone. And the calendar bites hardest in medicine: the last UCAT test day was 24 September 2026, ahead of a 15 October 2026 UCAS date, for courses where Oxford alone runs at 9.4 applicants per place. Around 37,000 candidates sit the test each year, and the ones who lose a cycle usually lose it to a booking deadline, not to a score.

How do you build a Year 13 plan around the collisions?

The useful thing about every date above is that it is published in advance, which makes the collisions plannable rather than discoverable. Lay them along one line and the year has four pinch points, and each one is the intersection of two things that both feel urgent.

The first is early autumn. The UAT-UK tests ran from 12 to 16 October 2026, inside the first half of the autumn term, while the first block of new A-level and IB content is still being taught for the first time. Test preparation therefore competes directly with the content that will be examined in June. Booking for those tests closed on 28 September 2026, before most students had a first assessment back, so the decision to sit had to be made without the evidence a student would normally use to make it.

The second is December. Oxford interviews are online in December 2026, with as little as a week's notice, and December is also mock season in most schools. The third is January: the 13 January 2027 equal consideration date lands immediately after those mocks, which means a student who has not drafted three personal statement answers by mid-December is writing them in the worst fortnight of the year. The fourth is June, where STEP sits inside the same exam season as A-levels, and where IB internal assessments have been running underneath everything else since the autumn.

None of that is an argument for working more hours. It is an argument for putting the fixed dates on a wall in September and deciding in advance which weeks belong to test preparation, which belong to the application, and which belong to the syllabus – then protecting each one from the others. The students who come unstuck in Year 13 are rarely the ones who worked too little. They are the ones who found out in October that something closed in September, and spent the rest of the year absorbing the cost of that.

Frequently Asked Questions