MMI circuit day: what actually happens, minute by minute

A multiple mini interview is a circuit. Several candidates rotate through the same set of rooms at once, one per station, and everybody moves on the same bell. Each room holds one task and one assessor, and the whole thing runs with the emotional warmth of an airport gate: numbers, doors, timers. That is not a flaw. The format exists because many short, separately marked samples of you predict performance better than one long conversation with a panel whose first impression colours everything after it.

Station content — role-play, data, ethics, motivation — is its own subject, and the MMI guide covers the station types in detail. This piece is about the machine around them: what the day does with you from the moment you check in, and how candidates who understand the machinery quietly outperform candidates who only rehearsed answers.

The thing nobody warns you about is the rhythm. There is no debrief between stations, no friendly reset, no moment where anyone asks how you are finding it. The bell is completely indifferent to whether you were mid-flow or drowning. Treat that indifference as a gift: it ends your worst station exactly as ruthlessly as your best one, and it does so every five to eight minutes for the whole lap.

Two caveats on the timings above. They are typical, not universal — some schools run longer stations, some run two circuits in parallel to get through the cohort, and online circuits compress differently. And the reading window varies more than any other number, so check your invitation for it specifically. Ninety seconds and two minutes feel like different sports.

Ninety seconds is enough time for orientation, not authorship. Candidates who spend the window drafting their opening paragraph walk in with a script, deliver it, and then collapse at the first follow-up question, because the script has no second page. The stronger move is to leave the door with a task, a role and a structure — the frameworks in structuring answers under pressure are built for exactly this window — and to compose the actual sentences in the room, where you can react to what the assessor or actor gives you.

Two reassurances. At many schools a copy of the prompt is available inside the station too, so you are not memorising against the clock. And if you finish reading with time spare, stop reading. Re-reading a prompt you have already understood is how a calm 90 seconds turns into a panicked one.

The walk between doors is when scoring happens. Each assessor sits with the same station all day, marking every candidate against the same rubric — usually a handful of station-specific criteria plus a global impression — and they settle your score in the gap while you are reading the next prompt. They write constantly during your station because writing is the job, not because you have said something alarming. Reading an assessor’s face is a mug’s game: the deadpan is trained neutrality, applied to every candidate, and treating it as feedback will wreck your delivery. What panels actually read in your body language, and what to do about it, is covered in nerves, delivery and presence.

What assessors do not do, in almost every UK circuit, is confer. Your ethics score is settled by the ethics assessor alone. Nobody at station six knows you stumbled at station three, and nobody aggregates anything until after you have gone home.

Online circuits reproduce the same machine in software. The bell becomes a chime or a countdown on screen, the prompt appears on a timed page, and an invigilator moves you between breakout rooms instead of corridors. The marking logic is identical: one assessor per station, scored independently.

Technology fails, and schools know it. As of the 2026 entry cycle, invitations to online interviews typically include a technical-support phone number, and interview platforms generally log disconnections with timestamps — so a dropped connection is usually visible to the admissions team, not your word against the void. If you lose the call: note the time, rejoin through your original link, and if you are not back inside a minute or two, ring the support number. Schools generally re-run an interrupted station or rebook the candidate where the fault was not theirs, but policies differ, so find your school’s stated process in the invitation before the day. What costs candidates is not the disconnection. It is the ten minutes spent frantically refreshing instead of following the recovery route that was in the email all along.

One last thing the timeline cannot convey: cumulative load. Almost anyone can be sharp for eight minutes. The circuit asks you to be adequate ten times in a row, while tired, with no feedback between attempts — and that is a stamina problem, not a knowledge problem. It trains quickly, but only under circuit conditions: full-length practice laps with a real timer and no pauses, not one polished station a night. Get two or three complete laps in before the real one and the day loses most of its strangeness. The bells stop being alarming and start being what they were designed as all along — a metronome, resetting your score to zero every eight minutes, all the way to the end.