OET Speaking Role Plays: How to Score 350+ in 2026

In short: OET Speaking is two profession-specific role plays of about five minutes each, and it is scored on nine criteria - four linguistic and five clinical communication. Candidates who reach 350 (grade B) rarely have the widest vocabulary; they follow the patient's cues, structure the consultation, and check understanding out loud. WispGold is an independent exam-preparation provider and is not affiliated with, endorsed by, or an official representative of OET. Confirm current formats and fees on the official OET website.

The first thing to understand is who you are talking to. The interlocutor plays a patient, a relative or a carer, and you play yourself in your own profession. You get a role play card with a setting, a task and cues. You have around three minutes to read it, and you are allowed to write on it. Most lost marks come from ignoring the card, not from grammar.

The linguistic criteria are intelligibility, fluency, appropriateness of language and resources of grammar and expression. The clinical communication criteria are indicators of relationship building, understanding the patient's perspective, providing structure, information gathering and information giving. Those five are where a competent clinician can gain marks quickly, because they reward the behaviours you already use at work when you slow down and say them aloud.

Relationship building starts in the first fifteen seconds. Introduce yourself with name and role, confirm who you are speaking to, and signal warmth without over-apologising. Then pick up the patient's stated worry and name it back to them. If they say they are frightened about the operation, do not move to the information task until you have acknowledged the fear in your own words.

Providing structure is the criterion most candidates neglect. Signpost the consultation: say what you will cover, move between sections explicitly, and summarise before you close. Phrases such as "I would like to do three things in the next few minutes", "before we talk about treatment, can I check what you already know", and "so, to summarise what we have agreed" are scored behaviours, not filler.

Information gathering means open questions first, then focused ones, and an explicit check that you have understood. Information giving means chunking - one idea at a time, plain words instead of medical terms, and a check-back after each chunk. Say "discharge home with support" rather than "disposition planning", then ask "does that sound manageable to you?"

The hardest cards are the ones where the patient refuses. A relative who will not accept a care plan, a patient who wants to leave against advice, a father who does not want antibiotics for his child. Marks come from exploring the reason for the refusal, respecting autonomy, offering a compromise within your professional limits, and leaving the door open. Never bulldoze the interlocutor to finish the task; an unfinished but well-handled negotiation scores better than a rushed lecture.

A four-week practice plan works well. Week one: record yourself doing two cards a day and listen back only for structure and check-backs. Week two: add a partner or tutor who plays the cues deliberately awkwardly. Week three: timed pairs of role plays back to back, so stamina and reset between cards become automatic. Week four: two full mock speaking tests with feedback mapped to the nine criteria, plus light polish on pronunciation and pace.

Practical exam-day points: speak at a slightly slower pace than normal conversation, keep your hands still, and do not read from the card once you begin. If you lose the thread, summarise - it buys time and scores a criterion. If your first role play went badly, treat the second as a separate exam; each is marked independently.

Most nurses and doctors who plateau at 300 are strong clinically and quiet about it. Say the caring out loud, structure the consultation, and check understanding twice, and the grade usually moves within a few weeks. Our OET course pairs weekly live role plays with criterion-mapped feedback so you can see exactly which of the nine areas is costing you marks.