MMI Practice Questions, Matched to Each UK Medical School's Published 2027 Format

Almost every MMI resource online is a list of scenarios. None of them tells you that Manchester runs 5 stations of 8 minutes while UCL runs up to 8 stations of 5, or that Leeds interviews on 4, 5 and 6 January 2027. This page attaches practice to published format, school by school, and says plainly where the evidence stops.

Which UK medical schools actually run an MMI for 2027 entry?

The first thing to settle is whether the school interviewing you uses a multiple mini interview at all. Several of the best-known medicine courses in the UK do not, and a candidate who spends December rehearsing station rotations for a course that runs a thirty-minute two-part panel has prepared for the wrong assessment altogether. Queen Mary University of London is the clearest case. Its published MBBS admissions policy declines to commit: the university "may offer either Multiple Mini Interviews (MMI) or panel interviews comprised of at least two individuals, all of which can be conducted either in person or virtually". That is a genuine either/or, and any resource listing Barts and The London as a settled MMI school is telling you more than the school itself has. Glasgow is definite in the other direction. For 2027 entry its medicine interviews are online, run to 30 minutes in total, and are split into Panel A and Panel B with two interviewers — a panel, not a circuit, and no amount of station practice changes that.

Oxford and Cambridge sit outside this conversation in a way the question banks never mention. Cambridge publishes a main interview period of 7 to 18 December 2026 and states that most applicants will have one or two interviews lasting a total of 35 minutes to an hour — academic interviews, not a circuit of scenarios. Oxford interviews online in December and is unusually blunt about the size of the pool: only around 425 applicants are shortlisted for interview each year, and on a three-year average covering 2023 to 2025, 29% of applicants were interviewed and 11% were successful. The Medical Schools Council puts the panel alternative at usually around 20 to 30 minutes long. If your five choices mix formats, you are preparing for two different assessments and should say so out loud when you build a practice schedule.

Read that as a practice plan rather than a league table. Nothing in it says one format is harder than another; it says the rehearsal has to differ. A Glasgow candidate needs to hold a coherent line for thirty minutes with two people watching. A Manchester candidate needs five self-contained eight-minute performances with a two-minute walk between them. A Barts candidate has to be ready for either, which in practice means preparing answers that work whether or not the person opposite has read the UCAS form. This is the same logic we apply across admissions tests: work out what the assessment actually measures and how it is administered before deciding what to drill. Question volume is the least useful variable in the whole exercise.

How many stations, how long, and what may you write down?

Timing is where published format changes practice most and where the SERP is emptiest. Manchester runs the shortest circuit of any major UK school — five stations — but the longest individual station, at eight minutes, with a two-minute gap between them. It also rules out anything on paper: "No information will be provided in advance, and there will be no reading or writing component to any of the online or in-person interview stations." That is a very specific rehearsal spec. It means eight minutes of sustained conversation, built live, with nothing to look at and nothing to scribble on. Practising Manchester with a printed scenario in front of you rehearses a skill you will not be allowed to use.

UCL sits at the opposite end. Up to eight stations, five minutes each, one minute outside the door to read basic instructions, and two or three questions inside each station. That is a different discipline entirely: convert a minute of reading into a structure, then deliver it in five while leaving room for two follow-ups. Leeds gives you more of both — six minutes inside the station and two minutes outside to prepare and make brief notes, with one interviewer per station. Birmingham breaks its stations into two minutes of preparation and six minutes of interview, moves candidates round a numbered circuit in a large room, and blows a whistle when it is time to move on.

Two schools publish what happens to your answer after you give it. Imperial marks each station out of six points for content, meaning what you say, and out of four for communication, meaning how you say it. That ratio is worth sitting with. A candidate who spends every rehearsal polishing delivery is optimising the smaller half of the mark. Bristol does something different again: six stations, each assessed by three assessors whose scores are combined, so you meet six assessors in total across two groups of three. A single unimpressed listener carries less weight there than in a Leeds station, where one interviewer is the whole record. None of this is a reason to change what you believe; it is a reason to change how long you talk, and to whom. Sixth-form students juggling this alongside A level coursework usually find that timing, not content, is what falls apart first under pressure.

When will your interview actually be?

The generic advice is that medical interviews run from November to February. For 2027 entry that window is too narrow at both ends, and building a practice schedule around it will leave some candidates peaking two months early and others caught short. UCL invites candidates "on a rolling basis from December through to the end of March" and makes the bulk of its offers towards the end of March. Glasgow runs from late November 2026 to March 2027. At the other extreme, Leeds has already published exact dates: face-to-face interviews for the MBChB on 4th, 5th and 6th January 2027, with overseas interviews carried out online. Three days, in the first week of January, immediately after the Christmas holidays.

Put your own five schools against that table and the shape of the autumn changes. A candidate holding Leeds and Cambridge has to be at full readiness by early December and effectively finished by the second week of January. A candidate holding UCL and Glasgow may still be interviewing in March, which is a very different problem: sustaining sharpness across three months without going stale, rather than compressing everything into a fortnight. Cardiff's deadline is worth noting even though it is not an interview date, because it is published to the minute — a fully completed UCAS application by 15 October 2026 at 18:00 UK time. Nothing about the interview matters if the application misses that. The practical move is to write the earliest published date among your five choices at the top of your revision plan and work backwards from it, rather than from the generic November-to-February line that most guidance still repeats.

What can the person in front of you actually see?

This is the detail that changes how you answer, and no page on the first results page for MMI practice questions raises it. At Newcastle, the 2027 admissions policy states that "Selectors will not have access to the applicant's personal statement or reference, either prior to and during an interview." At Sheffield, interviewers have no access to UCAS personal statements. Glasgow goes furthest: its own preparation advice tells candidates to revisit their personal statement precisely because "interviewers have no knowledge about you other than your name". Barts is the opposite. If you are shortlisted there, your interview panel "will have access to your UCAS application and may ask you questions from the information you have provided". Queen's University Belfast is more explicit still: the applicant's personal statement is considered within the MMI process itself.

Practise the wrong way round and you waste the opening of a five-minute station. A candidate who has rehearsed for Barts will open with "as I mentioned in my personal statement" and hand a Glasgow interviewer a reference to something they have never seen. A candidate who has rehearsed for Glasgow will spend forty-five seconds setting up a placement in front of a Barts panel that already read about it and wanted to probe it instead. The fix is not complicated: for the blind schools, every story needs a one-sentence self-contained opening — who, where, when, what you did — before the reflection starts. For the sighted schools, cut the setup and go straight to what you took from it, because the setup is already on the page in front of them. Rehearse both versions of your three or four strongest examples and label them by school. It takes an evening and it is the single highest-return preparation on this page, because it costs nothing in content and buys back time in the only currency the format cares about.

How much is the MMI actually worth?

Parents reasonably ask whether interview practice is worth the effort when grades and the UCAT have already been earned. The published answer varies more than most people expect, and at two schools it is total. Leeds states that "The decision to make an offer after interview depends entirely on performance at interview, not on predicted or achieved academic performance, or UCAT scores." Everything that got you shortlisted stops counting the moment you walk into the first station. Newcastle takes the middle position and publishes it precisely: the final score is based on the academic screen score and the interview score, weighted 50:50. Newcastle also folds the UCAT situational judgement band in as if it were an extra station, converting Band 1 to 5 points, Band 2 to 3 and Band 3 to 1. Sheffield does something structurally similar, converting the SJT to a score out of five and adding it as a virtual ninth station alongside the eight real ones.

Birmingham publishes the rule that should worry a strong candidate most. It sets a minimum performance standard for each station except the calculation stations, well below the average score, and states that "any applicant who does not reach the minimum standard for one or more stations may not receive an offer, irrespective of their overall interview score". That is a veto, not a weighting. It means the Birmingham candidate's job is not to be brilliant seven times; it is to be brilliant five times and never to be unrecognisably bad once. Most candidates who fall foul of a rule like that do so in the station type they quietly avoid in practice — usually role play, occasionally the numerical one. The honest implication is uncomfortable: rehearsing the stations you enjoy is close to worthless under a minimum-standard rule, and the whole return sits in the ones you have been putting off.

Data, numeracy and calculation stations: which schools publish one?

Three UK schools publish something recognisable as a data or numeracy station, and it is worth naming them precisely because so much advice names the wrong ones. Imperial lists "Data Interpretation" among the main topics its stations are drawn from, alongside teamwork and leadership, motivation to study medicine, understanding the role of a doctor, empathy and breaking bad news, ethics and values-related scenarios, and work experience questions — seven topics in all. Sheffield lists numeracy among its published station criteria. Birmingham is the most concrete: candidates undertake simple mathematical calculations involving data with clinical relevance, on a laptop provided at the station. Birmingham also tells you the level. "The mathematical skills that are being tested are at or below GCSE level, and those taking A level Mathematics are not necessarily at an advantage."

That last sentence deserves more attention than it gets. A station described as data interpretation sounds like it rewards mathematical firepower, and the school is saying plainly that it does not. What it rewards is doing arithmetic at GCSE level out loud, under time pressure, while a stranger watches and while you narrate what the number means clinically. Those are three separate skills and only one of them is maths. The way to practise is to take any published table of health data, give yourself the station's own time limit, and force yourself to do three things in order: say what the table is measuring and over what period, state the one comparison that matters and the figure that supports it, then say what you would want to know before drawing a conclusion. That third step is where most candidates stop early, and it is the one that separates a description from an interpretation.

Two cautions. First, resist the temptation to arrive with memorised statistics. Numbers move, and a confidently quoted figure that has since changed is worse than saying you would check the current published series. Second, remember Birmingham's exemption: the calculation stations sit outside its per-station minimum standard, so a shaky calculation cannot on its own veto an offer there, while a shaky role play can. Prioritise accordingly. If the school on your list runs no published data station at all, do not manufacture one; spend that evening on ethics or on the role play instead. Candidates coming through our subject tuition for sciences often have the arithmetic long since covered and need only the narration.

MMI practice for international applicants: a different format, not just a harder one

If you are applying as an international fee payer, the most important thing on this page is that the assessment you sit may not be the one described in any of the guides. Newcastle's 2027 admissions policy states that interviews for Home A100 and A101 applicants generally take the form of multiple mini interviews on campus, while "Applicants classed as International for fees purposes will generally be interviewed by a panel of two interviewers either on campus or remotely if this is more convenient". An international Newcastle applicant drilling MMI circuits is preparing for an assessment they will not sit. Birmingham keeps the MMI label but changes almost everything about it: for international applicants it runs two stations rather than seven, each lasting ten minutes made up of two minutes of preparation, six minutes of interview and two minutes of transition, on Zoom, with the whole interview no longer than thirty minutes. One of the two is a role play. Queen's University Belfast keeps the same MMI for both groups but delivers it online for international fee payers and in person in Belfast for home applicants.

Do the arithmetic on Birmingham and the strategic point is obvious. Seven stations forgive one weak performance; two do not, and one of the two is guaranteed to be a role play. An international Birmingham applicant should therefore spend a disproportionate share of preparation on role play, which is precisely the opposite of the "cover all eight themes evenly" advice the question banks give. On competition, the numbers that can be sourced are narrower than the ones usually quoted. Sheffield publishes its interview volumes: typically approximately 1,150 Home/EU applications and approximately 150 Overseas applications are invited to interview. Nationally, the Office for Students publishes maximum fundable limits for medicine in England, and those are the figures we now use on this page instead of the cap percentage the previous version carried.

Two things follow. The overseas share of fundable medical places in England is 5.6%, not the 7.5% figure repeated across the advice sites, and it is an England figure rather than a UK-wide one, so it says nothing directly about Scotland, Wales or Northern Ireland. And a number like that describes places, not your odds, because it says nothing about how many people apply for them. Where a school publishes both, as Sheffield does, use the school's own figures rather than a national average. The useful conclusion for an international candidate is not that the odds are long — you knew that — but that the format differs by fee status at named schools, and that the difference is published and checkable rather than a matter of rumour.

Why no question bank can contain a real MMI station

Every page competing for this search sells volume, promising question banks in the hundreds. It is worth being honest about what that volume can and cannot be. Before you attend a Leeds interview, you are asked to read and sign a non-disclosure agreement that sets out the rules you are expected to follow on confidentiality and professional standards of behaviour. Candidates who have sat the real stations are contractually and professionally barred from repeating them. So no published bank contains actual station content; it contains plausible reconstructions of station types. That is not a scandal, and the banks are not useless — but it does mean the value of practice questions lies in rehearsing a process, not in acquiring content you might see again.

There is one published exception, and it proves the rule rather than breaking it. Sheffield sends candidates the themes of its MMI stations when confirming the interview booking. Not the questions: the themes. A school willing to tell you the subject matter in advance is telling you something about what it is assessing — if knowing the theme were enough to score well, there would be no point sending it. What Sheffield is measuring is what you do inside eight minutes with a theme you have had time to think about, which is exactly the skill an unseen station measures too, minus the surprise.

Treat this as liberating rather than discouraging. It means the return on grinding through a fourth list of 300 scenarios is close to zero once you have covered the recognised themes, and the return on rehearsing under the real constraints — the real clock, the real reading time, a real stranger, no notes if your school forbids them — is high and keeps compounding. It also means the honest test of readiness is not how many questions you have seen. It is whether you can take a scenario you have never met, in the format your school publishes, and still produce something structured. That is the same argument we make about sixth-form entrance interviews: familiarity with the format beats familiarity with the questions, every time.

Building a practice circuit that matches your schools

Here is how to turn everything above into a fortnight of useful work. Start by writing your five schools down the left of a page and, next to each, the three facts that govern rehearsal: how many stations, how long each one runs, and whether there is reading time. From this page that is quick — Manchester 5 and 8 minutes with nothing to read; UCL up to 8 and 5 minutes with a minute outside; Leeds 8 and 6 minutes with 2 minutes outside; Birmingham 7 with 2 minutes of preparation and 6 of interview; Sheffield 8 at typically 8 minutes. Then add the fourth column that almost nobody fills in: what the interviewer can see. If two of your schools sit at opposite ends of that column, you have two versions of every answer to prepare, and you should know that in November rather than discovering it in the room.

Then rehearse in circuits, not in questions. A single practice session should run the number of stations your school publishes, back to back, with the real gaps — two minutes at Manchester, one minute of reading at UCL — and no stopping to discuss between stations. The point is to experience what a poor station does to the next one, because that is the failure mode nobody rehearses. Feedback comes at the end, on the whole circuit. Only when a whole circuit holds together is it worth going back to isolated station types. If time is short, spend it on role play and on whichever station type you have been avoiding, because under a minimum-standard rule like Birmingham's the downside sits there and nowhere else.

Finally, keep your dates honest. Leeds interviews on 4, 5 and 6 January 2027, Cambridge between 7 and 18 December 2026, Nottingham's A100 from 15 December 2026 to 29 January 2027. Those are published, and they mean the practice window for many candidates closes before the spring term properly starts. If your earliest school is Leeds, a plan that has you peaking in February is a plan that has already failed. Working backwards from a real published date is unglamorous, but it is the difference between a candidate who is ready and one who was going to be.

Frequently Asked Questions