Five tasks, one microphone, and a three-second rule.
PTE Speaking is scored by a machine listening for three things: content, oral fluency and pronunciation. That changes the strategy completely. You are not charming an examiner; you are feeding a scoring engine that rewards steadiness, penalises silence, and never gives credit for a beautiful phrase it could not parse.
Why this matters. The single most expensive habit in PTE Speaking is the natural one: pausing to think. A silence of around three seconds closes the microphone, and everything unsaid scores nothing. The second most expensive is racing, because fluency is scored on rhythm and stability, not speed. Steady, continuous, clearly stressed speech at a conversational pace is the whole game.
What each one asks, and what it really tests.
An unscored personal introduction opens the test. Then these five, in order.
Read Aloud
Feeds Reading + Speaking
A short text on screen; read it aloud after the tone. Preview time is for scanning stress and chunking, not for understanding every word.
Practise by marking the two or three stress peaks in each sentence first.
Repeat Sentence
Feeds Listening + Speaking
Hear a sentence once, repeat it. Hold the meaning, not the string: candidates who chase the exact words drop the second half; candidates who hold the idea usually reproduce most of it.
Say something continuous even when memory fails. Partial credit is real.
Describe Image
Speaking only
Twenty-five seconds to look, forty to speak. A fixed frame works here and is not penalised, because the content is the chart in front of you: what it shows, the extreme values, one trend, one conclusion.
Overall → highest → lowest → trend → conclusion. Every image, same route.
Re-tell Lecture
Feeds Listening + Speaking
Notes while you listen, then a steady spoken summary. Three or four content points delivered fluently beat eight fragments.
The lecture discussed… the speaker argued… finally, they concluded…
Answer Short Question
One word or a few
General-knowledge vocabulary questions. Answer immediately with the word; do not build a sentence around it.
What do you call the doctor who treats teeth? A dentist. Done.
Three traits, and the discipline each demands.
Every speaking response is scored on some combination of these. None of them is impressed by rare vocabulary.
| Trait | What it measures | The discipline |
|---|---|---|
| Content | Did you say the things the item asked for | Fixed routes: the image route, the lecture frame, the held meaning of a sentence. |
| Oral fluency | Rhythm, continuity, stability of pace | Never stop. A wrong word delivered smoothly outscores a long correcting pause. No um, no er, no restarts. |
| Pronunciation | Clear sounds, sentence stress, intelligibility | Accent is not the issue; clarity is. Stress the content words, let the grammar words stay light. |
The three-second rule, plainly. If you stop for around three seconds, the microphone closes and the item ends. So the one unbreakable habit is: keep producing connected speech until the time runs out or you finish. Trained steadiness beats natural brilliance in this room, every time.
Short, daily, out loud.
Speaking practice for a machine-scored test is repetition with feedback, not conversation. Three drills carry most of the value.
Read aloud daily
Two minutes, any academic text
Mark the stress peaks, then read for rhythm. Record yourself once a week and listen for hesitations, not accent.
This one drill feeds Reading and Speaking at once.
Shadow sentences
Repeat-Sentence training
Play a sentence, hold the meaning, say it back without a pause. Do it in tens, daily; the length you can hold grows with practice.
Start at eight words; twelve to fourteen is exam-comfortable.
One image a day
The fixed route, timed
Twenty-five seconds looking, forty speaking, the same route every time. The route is the fluency; the chart merely supplies the nouns.
Overall → highest → lowest → trend → conclusion.
Where a tutor actually helps
Self-practice builds steadiness, but it cannot hear your own sentence stress from the outside. In a lesson I listen the way the scoring engine does, for continuity, stress and clarity, tell you precisely which habit is costing you, and drill the repair. As a London-born speaker of Received Pronunciation, the model you copy in a lesson is the clean, clearly stressed English the machine parses best.
Steadiness is a habit. Habits are trainable.
The machine rewards exactly what disciplined practice builds: continuous, clearly stressed, unhurried speech.
Drill the three exercises above daily, and read the rest of the section: Writing, Reading, Listening. For feedback a recording cannot give you, lessons are £20 for fifty minutes, one to one, in proper British English; the first step is a discounted trial lesson.