How long is OET Speaking? 20 minutes, timed

Most candidates who score Grade C on OET Speaking do not have weak English. They have weak pacing. They spend two minutes on the opening, run out of time to demonstrate half the rubric criteria, and leave the booth feeling the role-play went well only to see a report full of "insufficient evidence" flags for behaviours they simply never got to.

How long is the OET Speaking sub-test?

About 20 minutes, and only ten of them are scored:

Allow 20–30 minutes at the venue once identity and security checks are counted. The warm-up exists to settle your nerves and is deliberately not assessed — everything from the first line of role-play 1 onward is.

The marks are distributed across the whole five-minute span of each role-play, so spending too long on one section does not only cost you that section: the sections you never reach score nothing at all. The rest of this post is the minute-by-minute breakdown of the three preparation minutes and the five role-play minutes, tuned for a candidate sitting the exam within the month. For the full structure of the sub-test, see the OET Speaking sub-test guide.

If you have not read the underlying scoring model yet, how OET Speaking is actually scored covers the nine criteria the examiner is watching for. This post assumes you know the criteria; it is about the when and the how long, not the what.

The three preparation minutes

The interlocutor hands you the role-play card and you get three minutes with it before the role-play begins. You may take notes on the noteboard provided, and you may ask the interlocutor about any word on the card you do not recognise — a right most candidates never use. What most candidates do instead is read the card four times, panic-plan a script, and step into the role-play with a rigid mental outline that falls apart the moment the simulated patient says something unexpected.

Those three minutes are not for scripting. They are for mapping. Spend them on four things and nothing else:

• Read the card once, at reading speed (about 45 seconds). Get the role, the setting, and the four to five task points into your head as bullets, not as sentences.
• Ask about anything you cannot read (about 15 seconds, often zero). If a drug name, condition or piece of equipment on the card is unfamiliar, ask the interlocutor now. It costs nothing and it is far cheaper than stumbling over it mid-role-play.
• Note the two hardest tasks (about 45 seconds). One is usually explaining something clinically technical in plain language; another is usually managing an emotional response or a piece of resistance. Those are the tasks that burn time if you improvise, so know they are coming and jot the two or three words you will reach for.
• Decide your opening line (about 30 seconds). Not the whole opening — just the first sentence. Something like: "Good morning, I'm one of the nurses on the ward — is now a good time for us to talk about your discharge plan?" Having the first sentence pre-loaded means you do not stall in the first ten seconds of the recording.

That leaves roughly 45 seconds of contingency. Do not fill it by re-reading the card or silently rehearsing. Sit still, breathe out, and let the clock run down. Candidates who spend all three minutes in high-effort mental preparation walk into the role-play already cognitively tired, and the role-play is where every mark actually is — arrive with capacity.

The 5-minute role-play arc

The role-play unfolds in four phases, and the marks are distributed unevenly across them. If you spend equal time on each, you starve the middle phases — which is where most of the rubric evidence lives.

A rough budget for a five-minute role-play (300 seconds):

• Phase 1 — Opening / rapport (30–45 seconds)
• Phase 2 — Information gathering (60–90 seconds)
• Phase 3 — Information giving / explanation (90–120 seconds)
• Phase 4 — Negotiation, resistance handling, closing (60–75 seconds)

Those numbers add to roughly 240–330 seconds. The bracket is deliberately loose because different task cards weight the phases differently. A discharge-planning role-play spends more time in phase 4; a diagnostic explanation role-play spends more in phase 3. The important thing is that no phase is skipped, because each phase is where distinct rubric criteria collect their evidence.

Phase 1 — Opening (0:00–0:45)

The first thirty to forty-five seconds establish the interaction and score the initial Relationship-Building marks. What you must do here:

• Greet, identify yourself and your role clearly.
• Check that now is a good time and briefly frame what you are here for.
• Acknowledge the patient as a person before you start any clinical exchange.

That is it. Do not begin explaining the clinical issue yet. Candidates who charge straight into content in the first 20 seconds save time on paper but leave the Relationship-Building criterion under-evidenced, and the examiner cannot retroactively award marks for rapport you skipped.

A common mistake: over-doing the opening for two full minutes because it feels safe. The rapport-building marks max out by the 45-second mark. Beyond that, every extra second of pleasantries is a second stolen from the middle phases where more marks live.

Phase 2 — Information gathering (0:45–2:15)

This is where "finding out and understanding the patient's needs" scores. You have about 90 seconds. Two things must happen:

• Ask at least one broad open question early, then let the patient run for 20–30 seconds without narrowing. "Tell me a bit about how things have been at home since your last visit" gives the examiner a broad open-question evidence anchor.
• Follow up with two or three probing questions that visibly track what the patient said. Reflecting back specifically — "you mentioned the stairs have been an issue — how are you managing those right now?" — earns the criterion in a way that a fresh, unrelated question does not.

Candidates who blow phase 2 usually do so by asking a rapid-fire closed-question sequence — "Have you been taking the medication? Any side effects? How's your pain?" — that gets to the clinical answer in 45 seconds. Efficient, and rubric-poor. See what makes OET Speaking hard for the deeper structural reason this failure pattern is so common in clinicians.

If you find yourself already talking about your own recommendation before the 2-minute mark, you are ahead of schedule and phase 2 is under-developed. Slow down.

Phase 3 — Information giving (2:15–4:15)

Roughly two minutes for the explaining-and-checking phase. This is where two of the biggest-weight criteria — appropriateness of language for a layperson, and providing structure — collect most of their evidence.

The disciplined rhythm here is chunk → check → chunk → check. One idea, comprehension check, next idea, comprehension check. Not "here is everything you need to know about your discharge plan, does that all make sense?" — that is one giant chunk with one weak check, and it scores as one weak check. Four small chunks with four checks is the target.

A worked five-chunk example, delivered over 90–100 seconds:

1. "The main thing we want to talk about today is what happens when you leave the hospital. Does that sound OK?" (10 sec — frame + check)
2. "Your medication will change slightly — you'll be on the same tablets but a lower dose. Does that make sense?" (15 sec — content chunk + check)
3. "We'd like a nurse to come and see you at home twice in the first week — is that something you'd be comfortable with?" (15 sec — content chunk + check with negotiation flavour)
4. "There are a few things to watch out for — a temperature over 38, or the wound looking red or hot. Have you come across those signs before?" (20 sec — content chunk + check + optional teach-back)
5. "I'll write all of this down so you don't have to remember it. Anything so far you'd like me to go over again?" (15 sec — reassurance + open-check invitation)

Total: about 75 seconds. That leaves you 15–20 seconds of headroom in phase 3 for handling whatever the patient says back to those checks — because they will say something, and following up on it is where the marks fatten.

Jargon glossing lives inside these chunks. Every clinical term gets an inline plain-language gloss the first time it appears: "your renal function — that's how well your kidneys are working." See OET Speaking for doctors for the specific version of this trap that hits medical practitioners hardest.

Phase 4 — Negotiation, resistance, closing (4:15–5:00)

The last 45–60 seconds are where the shared-decision-making and empathy criteria have their strongest evidence window. Almost every OET role-play card includes a task point that involves the patient objecting, hesitating, or expressing a fear — the card is designed that way, because the rubric wants to see negotiation and empathy under pressure.

If you have paced phase 3 correctly, you enter phase 4 with the patient partly on board and one or two objections still to resolve. What you do here:

• Voice the empathy explicitly when the objection lands. "I can hear that's really worrying you." Not implicit warmth, not a soothing tone — an audible sentence that names the emotion. See OET Speaking under exam stress for why this feels performative and why it still has to be done aloud.
• Offer options, not a single directive. "One option is that we could try the home visits and see how you go; another is that we could arrange a phone call first if that feels less intrusive. What do you think would work better for you?"
• Confirm the plan and close warmly. Summarise what was agreed in one or two sentences and offer a next-step contact. "So we'll arrange the two home visits and I'll pop back in on Thursday morning to see how you're settling — sound good?"

That is a natural close and it fits inside 45 seconds. Candidates who arrive at phase 4 with 90 seconds remaining tend to over-extend the closing and drift into repetition; candidates who arrive with 15 seconds remaining truncate the negotiation entirely and lose the criterion. Getting the pacing right in phases 2 and 3 is what protects phase 4.

What to cut when you are falling behind

Every candidate has at least one role-play where the patient goes off on an unexpected tangent and the clock accelerates. Two rules for recovery:

• Never cut Phase 4. If you are at 3:30 and still deep in explanation, cut the current chunk, do one visible comprehension check to close it, and pivot: "There's a bit more we could discuss, but I want to make sure we talk about how you're feeling about the plan overall." That transition itself demonstrates rubric-relevant behaviour (patient perspective, providing structure) and buys you the negotiation window.
• Cut chunks from the middle of Phase 3, not the end. Losing the last chunk deletes your last comprehension check; losing a middle chunk just reduces the density of content. The examiner is scoring the presence of the check-rhythm, not the completeness of the clinical content.

The one thing to never do is race through the ending to fit everything in. A rushed final 30 seconds with three chunks and no checks scores worse than a calm 30 seconds with two chunks and full checks.

How to practise timing at home

Timing intuition does not develop by reading about it. Three drills:

• Silent countdown drill (2 weeks before the exam): Set a phone timer for 5:00 face-down. Do the role-play out loud without looking. Turn the timer over the moment you finish. Repeat until your finish-time consistently lands between 4:45 and 5:00. This trains the internal clock so you do not need to visually check.

• Phase-tag drill (1 week before the exam): Do a role-play and record yourself. Play it back with pen and paper, marking the timestamp where each phase transition happened. Compare to the target windows above. Adjust one phase's tempo per drill — trying to fix all four at once fails.

• Interruption-recovery drill (final week): Have a practice partner (or an AI patient) deliberately derail the role-play at the 2-minute mark with a long, off-topic tangent. Practise recognising it happened, pivoting cleanly, and still landing all four phases inside 5 minutes. This trains the calibration you actually need on exam day.

The 8-week preparation roadmap has the timing drills sequenced against the broader skill work; run them alongside, not instead of, the rubric-behaviour drills.

The three timing traps

Certain patterns show up in almost every set of practice recordings a candidate submits:

Trap 1: over-invested opening. Two minutes of "how are you feeling today, is now a good time, before we start I just want to check…" burns half the exam. The Relationship-Building marks were secured by 0:45. Move on.

Trap 2: monologue phase 3. Forty-five uninterrupted seconds of clinical explanation with a single "does that make sense?" at the end. That is one weak check, not five strong ones, and the rubric's structure criterion scores accordingly.

Trap 3: skipped negotiation. Phase 4 gets swallowed by phase 3 overrun, the patient's objection lands with 20 seconds left, and the candidate wraps up without genuinely engaging it. The empathy and shared-decision criteria then score on almost no evidence. This is the single most expensive timing mistake because it wipes out two of the highest-weight criteria at once.

If you record five practice role-plays and one of these patterns shows up in three or more of them, that is your practice priority. Fix the pacing before you fix anything else — most of the "content" fixes you would otherwise be tempted to make are downstream of a timing problem.

Exam-day mechanics

A few things that trip candidates up on the actual day, unrelated to English:

• Preparation starts silently. You will not get a countdown or a warning partway through. The interlocutor hands you the card and starts the clock. Assume it is running the moment the card is in your hand.
• Neither the examiner nor the patient will slow down to help you catch up. If you fall behind, you catch up alone.
• The recording stops at 5:00 sharp. No fade-out, no "one more sentence" grace. Mid-sentence at 4:59, cut. This is why the interruption-recovery drill matters — you need to know where you are in the arc without checking a clock, because there is no clock to check inside the booth.

The exam-day version of you needs to be pre-boring

The most useful mental model for exam-day timing: your goal is to be slightly boring. Slightly slower than feels natural, slightly more structured than feels natural, slightly more explicit about empathy and negotiation than feels natural. The rubric rewards the deliberate, unhurried, ritualistic version of clinical communication — not the fluent, efficient, real-ward version. If you leave the booth feeling that the interaction was a little stiff and slow, you probably paced it right.

Candidates who leave the booth feeling they had a "natural, quick, warm chat with the patient" are almost always the ones who under-evidence half the rubric. The exam booth is not a real ward. Give it the ritual it wants for five minutes.

Practise the timing. The English is usually not the problem.