Interview questions with model answers, written by specialist subject tutors.
Panel interviews, not MMI stations: two conversations of twenty to thirty minutes with two or three Fellows, one of them a practising clinician, and something in front of you that you have not read before. One such hand-over — two blood results, nothing else — is reasoned out below in full, first attempt included, and the first attempt does not hold.
Medicine interviews at Oxford and Cambridge test scientific reasoning under a very specific kind of pressure: no right answer to recall, only a scenario to reason through aloud while two or three Fellows watch. The Medicine pack gives you ten of those scenarios — mole and concentration chemistry, pedigree-based inheritance, kidney and fetal-circulation physiology, and ethics on euthanasia and confidentiality — argued out in full, not left as a one-line answer key.
The Medicine pack — £180
Ten questions across 22 pages, each set out in three stages: the question alone, the hints a panel would offer once you stall, then the finished answer. One PDF, one payment, instant download.
Get the Medicine pack — £180Read a free sample firstThere is no Medicine sample. The nine free samples are other subjects, but every pack is built the same way — the question on its own, then the prompts, then the worked answer — so any of them shows what you would be getting.
Buying more than one? 10% off 2 packs · 20% off 3 or more. Applied automatically at checkout — there is no code to enter.
Straight to payment — no account, instant download.
Oxford and Cambridge Medicine interviews do not use the MMI (Multiple Mini Interview) format. Both universities use traditional panel interviews — typically two per candidate, each 25–30 minutes, with two or three Fellows including at least one clinician. You will be given a scenario, graph, or question and asked to reason through it aloud. There are no right answers to memorise. The assessment is entirely about whether your thinking process reflects the kind of mind that medicine at Oxford or Cambridge is designed to train. One hand-over of that kind is worked below from the bare question to a finished answer, with each nudge a panel would give set down at the point it would land; the Medicine pack works the genetics, public-health and ethics questions to that same depth, for £180.
Shortlisted candidates at Oxford are typically called to their applied college for two panel interviews. A panel includes two or three Fellows, at least one of whom is a practising clinician. The format replicates the undergraduate supervision — a sustained intellectual conversation in which you work through unfamiliar material under observation. Cambridge Medicine interviews follow an almost identical structure. Neither university scores interviews on a station rubric; the entire conversation contributes to the assessment. Approximately 150 students are admitted to Oxford Medicine annually and 280 to Cambridge, from a combined applicant pool of around 3,000 at each university.
| Factor | Oxford Medicine | Cambridge Medicine |
|---|---|---|
| Interview format | Panel (not MMI) | Panel (not MMI) |
| Interviews per candidate | 2 at applied college | 2; pool interview possible |
| Duration each | 25–30 minutes | 20–30 minutes |
| Pre-interview test | UCAT | UCAT (cognitive score only) |
| Annual intake | ~150 | ~280 |
| Typical offer rate from interview | ~25–30% | ~30–35% |
Scientific reasoning questions. You will be given data — a graph, a table, a brief experimental description — and asked to interpret it. "What does this graph show?" "What would happen to this value if the dose doubled?" The key is not knowing the answer in advance but reading the data carefully and reasoning from biological first principles. Oxford Medicine interviews are particularly well known for this question type, and Cambridge panels lean on it just as hard: with no written test before the interview and no submitted work, the reasoning has to happen out loud in the room.
Ethical scenario questions. Both universities test ethical reasoning. Oxford tends to use scenarios that require applying the four-principle framework (autonomy, beneficence, non-maleficence, justice) without the framework being named. Cambridge ethics questions tend to be more socially complex: resource allocation, the ethics of AI in healthcare, health inequalities. In both cases, reaching a verdict too quickly is the most common mistake — interviewers reward structured consideration of all affected parties over confident but shallow conclusions.
Personal statement deep-dives. Any claim in your personal statement is fair game. If you mentioned a specific book, you will be asked for its argument and whether you agree. If you mentioned work experience, you will be asked what you observed and what it made you think about medicine. Specific, honest, reflective answers score well; generic claims about "developing communication skills" do not.
“This patient’s plasma creatinine was 70 micromoles per litre when it was last measured, two years ago. Today it is 140. Neither reading was rising at the time it was taken. What has happened to their kidneys?”
Nothing else is supplied: no age, no symptoms, no scan, no reference range. That is deliberate. The panel is not checking whether you have read about kidney disease, but what you do with a proportionality when it is the only thing on the table.
The nudges an interviewer uses when the room goes quiet, in roughly this order:
The version I had to abandon. My first move was to treat creatinine as a damage meter. It has doubled, so the kidneys are half as good as they were — and then, because a rise of 35 is a rise of 35, a further climb to 175 would cost about as much again and leave roughly a quarter of the original function. The first half of that survived. The second half was a straight line drawn through two points, and the interviewer only had to ask me to test it at a third.
Starting again from where the prompt pointed. Creatinine is made at a roughly steady rate by muscle and cleared almost entirely by filtration at the kidney. If the measured level is not moving, then those two rates are equal: production = filtration rate × concentration. Rearranged, filtration rate = production ÷ concentration. With production held fixed, filtration rate is inversely proportional to the number the laboratory prints, and nothing else in the relationship is doing any work.
Call the original filtration rate G. At a concentration of 70 it was G. At 140 the concentration has doubled, so the rate is 70 ÷ 140 = one half of G. Half of this patient’s filtration is gone — the part of my first answer that held up.
Now the second patient, the one who went from 70 to 105. That is 70 ÷ 105 = 0.67, so they are filtering at about two thirds of their own baseline: a loss of a third. Half the rise on the page, but two thirds of the damage the first patient has. Turn it round and the same arithmetic says something sharper still. The first 35 points of rise, 70 to 105, cost 33% of baseline function. The next 35 points, 105 to 140, cost only a further 17% — the gap between 0.67 and 0.5. Identical movement in the printed number, twice the meaning the earlier it happens.
Which is where my straight line died. Extrapolating linearly through the two points I had been given predicts 25% of baseline at a concentration of 175. The relationship actually gives 70 ÷ 175 = 0.4, so 40%. My line understated what was left by fifteen points of the original function, and it erred in the direction that makes a bad result look worse than it already is. Run the curve the other way and it is the early end that should worry you: a patient who has lost a fifth of their filtration sits at 70 ÷ 0.8 = 87.5, a rise of seventeen and a half points. Why early loss is hard to see in this one number is now an argument you can make from the algebra rather than remember from a textbook.
What would make me wrong. All of it rests on production being constant, and that is the assumption I would put on the table before anyone asked me for it. Creatinine comes from muscle. A patient who has lost muscle over those two years — through illness, through immobility, through age — makes less of it, and a falling production rate holds the concentration steady while the filtration rate drops underneath it. On this reasoning a stable creatinine is not the same thing as a stable kidney, and the patient I would trust my own answer least about is the frail one.
Three things, and none of them is knowledge about kidneys. Whether you will convert a sentence about a patient into a relationship you can actually manipulate: the entire question turns on writing production = rate × concentration and then dividing by something. Whether you notice when you have drawn a straight line through a curve, which is the commonest way a confident candidate goes wrong on data handed to them cold. And whether the assumption holding your answer up is one you volunteer or one they have to dig out of you — the credit is not for being right about muscle mass, it is for having gone looking for the thing that could break your own argument before anybody else did.
The questions are the cheap part.
Oxbridge Medicine questions are all over the internet, this one among them. What is not is the nudge that arrives after the silence, and an answer written at the length you would have to speak it. The Medicine pack sets its questions with both of those attached and held back from the question itself, across genetics and inheritance argued from a pedigree diagram, public-health policy, ethical scenarios taken through the four principles, and the work-experience and motivation questions. Panel questions throughout, since that is what Oxford and Cambridge set. A PDF you can print, written by specialist subject tutors.
Medicine opens no preview file, so I am not going to quote you a page count for it. The creatinine question above is the test instead: if the reasoning in it was worth your evening, the pack is more evenings of the same thing. One payment of £180.
Get the Medicine pack — £180 →Some subjects on the pack page hand you a free PDF before you pay. Medicine does not, and that is better said here than discovered at the basket. Two things follow. The creatinine question above is the only whole thing on this subject you can have for nothing, which is why it runs all the way to a finished answer instead of stopping at an illustration. And if you want a real pack question in your hands before spending anything, open the Vet Med sample: two questions, the interviewer’s hints on a page of their own, and both answers worked out. It is the closest thing we give away to a clinical science file, and it will show you how long an answer runs and how much of the work is deliberately left to you.
The most common mistake is reaching a conclusion too quickly. Interviewers are looking for a candidate who identifies all relevant considerations, assigns appropriate weight to each, and arrives at a defensible position while acknowledging competing interests. Effective structure: (1) identify who is affected and what interests they have; (2) identify which ethical principles apply and whether they conflict; (3) note what additional information would change the analysis; (4) offer a tentative position with explicit reasoning. Our Oxford Medicine interview guide shows this structure applied to real scenarios.
UCAT is used by both Oxford and Cambridge to decide who is called to interview, and neither publishes a cut-off. Oxford states it sets no fixed minimum; Cambridge states that there is no minimum threshold for UCAT score, and for 2027 entry reads the overall cognitive subtest score with the situational judgement section left out. Any specific safe-score number you have been given for Oxbridge is somebody’s estimate rather than a published bar, so treat it as one. Neither university sets a written test before the interview, and Cambridge asks for no submitted work at all. Sit the UCAT in the summer window — booking for 2026 closes on 16 September — then move to interview preparation, because once you are shortlisted the conversation is what decides the offer.
The most effective preparation is working through unseen material — scenarios, data, and questions you have not encountered — with someone providing real-time feedback on your reasoning process. Knowing standard question answers in advance is far less useful than developing the instinct to structure your thinking explicitly and communicate it clearly under pressure. For scientific reasoning, practise interpreting graphs from clinical journals. For ethics, apply the four principles to a new scenario each week. For personal statement questions, re-read your statement and prepare substantive answers to follow-up questions on every claim you made. Whatever you practise on, practise it out loud and in front of somebody: reading an answer silently and delivering it while being interrupted are different skills, and only the second one is what the panel sees.
“My panel at Gonville & Caius handed me a short article about a clinical trial and asked what I thought the key limitation was. I’d never seen the paper before. The pack was the only preparation I found that actually trains you for that — reading through the model answers showed me how to reason about evidence out loud, identifying what is missing or uncertain rather than just summarising what is there. By the time I got into the room I knew how to think, not just what to say.”— Priya S., Medicine, Gonville & Caius Cambridge, 2024 entry
"My tutor at Balliol pushed back on everything I said. Every time I made a point, he'd say 'but surely...' and take the opposite position. I wasn't expecting that at all. The pack was the only resource I found that actually prepares you for that — the model answers show you how to structure an argument and defend it under pressure, not just state a view. Really glad I used it."— Ella T., History, Balliol College Oxford, 2025 entry
Oxford and Cambridge both use traditional panel interviews, not MMI stations. Candidates typically have two interviews of 25–30 minutes each, with a panel of two or three Fellows including at least one clinician. You will be presented with scenarios, data, or questions and asked to reason through them aloud. There are no scored stations, no time pressure, and no role-play elements. The format resembles an undergraduate supervision — a sustained intellectual conversation — which is why interviewers value sustained, structured reasoning over quick confident answers.
UCAT is used by both Oxford and Cambridge to help decide who is called to interview, but neither publishes a threshold. Oxford sets no fixed minimum cutoff, and Cambridge states that there is no minimum threshold for UCAT score, reading the overall cognitive subtest score and leaving the situational judgement section out for 2027 entry. Cambridge sets no written test before the interview and asks for no submitted work. Sit the UCAT in the summer window, then move to interview preparation: once you are shortlisted, the conversation is what decides the offer.
Ethical questions typically involve resource allocation, patient autonomy, end-of-life care, public health policy, and the ethics of medical technology. Questions rarely have a single correct answer. Interviewers expect you to identify all relevant considerations before reaching a position, applying the four principles of medical ethics — autonomy, beneficence, non-maleficence, and justice — without necessarily naming them. Practise with real ethical scenarios, focusing on the structure of your reasoning rather than memorising conclusions. Reaching a verdict too quickly is the most common and penalised mistake.
Work experience matters because it is the source of specific examples during the interview. Interviewers will follow up on anything you mentioned in your personal statement: what did you observe in a clinical setting, what surprised you, what changed your understanding of medicine as a practice. Generic answers ('I developed communication skills') score poorly. Specific, reflective observations from real clinical placements score well. Aim for at least 70–100 hours of clinical exposure, ideally in settings where you observe clinician-patient interactions rather than administrative work only.
Scientific reasoning questions present you with data — a graph, a table, a brief experimental description — and ask you to interpret it without prior knowledge of the subject. Common types include interpreting dose-response curves, reasoning about what a physiological measurement implies for organ function, and applying first-principles biology or chemistry to a clinical context. The assessed skill is methodical analysis: read the axes carefully, identify what the data shows, reason about what it implies, and identify what additional data would confirm or refute your interpretation.
No, and if it were it would not be worth paying for — Oxbridge Medicine questions are freely available, this one included. What the pack adds to every question it sets is the follow-up that comes when you stop talking, and then the answer at full length, both held on pages of their own away from the question, so that you meet each question the way you will meet it in December: cold, with nothing underneath it. The questions cover genetics and inheritance argued from a pedigree diagram, public-health policy, ethical scenarios taken through the four principles, and the work-experience and motivation questions, all in the panel format Oxford and Cambridge use rather than in MMI stations. It is a PDF you can print. Written by specialist subject tutors, £180 paid once, and Medicine opens no preview file, so the creatinine question on this page is the specimen.
How you behave when you are told you are wrong. That is the specific gap, and in Medicine it costs more than the speaking practice everybody worries about. Alone with a written answer you are never contradicted, so you never find out which way you go when a Fellow tells you flatly that the kidneys are simply half as good as they were, and then waits. Two ways are open. One is to hold the position and give the reason again more carefully, which is what the panel is hoping to see. The other is to hear an instruction in the question and drop a view you had good grounds for, which reads across the table as never having held one. Nothing printed can tempt you into the second, so nothing printed can train you out of it. What print does do is make sure there was a real position there to defend, and that half you can settle before December.
Further Reading: For Oxford Medicine interview questions with full worked answers on scientific reasoning and ethical scenarios, see our companion guide: Oxford Medicine Interview Questions 2026 — With Model Answers.
Rehearse the retraction, not just the answer
The worked question above contains an attempt that fails and a sentence in which it is given up. That sentence is the one almost nobody has practised. Candidates arrive knowing what they think; very few have any idea how they will sound in the ten seconds after a Fellow has shown them that what they think does not hold, and those ten seconds are being watched closely. Say it aloud now, in the abstract, until the words are simply available to you: that was wrong, here is the step that broke, here is what I would put in its place. Then get more questions to keep doing it on.
Genetics and inheritance argued from a pedigree diagram, public-health policy, ethical scenarios taken through the four principles, and the work-experience and motivation questions you will actually be asked. Every question carries the interviewer’s follow-ups and a full worked answer, both set apart from the question so you can attempt it cold. Panel questions throughout. A PDF you can print, written by specialist subject tutors, £180 paid once.
Trustpilot rates Leading Tuition Excellent, 4.8 out of 5 — a score for the tuition, earned by tutors, not a verdict on this file. Medicine opens no preview, so the creatinine question above is the writing you would be buying. If it did not convince you, do not buy the pack. That is the company’s rating, not a rating of this pack.
Get the Medicine pack — £180