ScoreFluentThese are the PTE speaking questions students ask most, with plain answers. They come from YouTube comments, Reddit threads, app-store reviews and Quora threads we read in October 2026, with wording lightly edited for clarity. Where Pearson has published a rule, we cite it. Where it has not, we say so.
Key facts
- PTE Academic is scored 10–90 by an AI scoring system. Some responses are also reviewed by a human expert (Pearson scoring page, read October 2026).
- Most speaking tasks are scored on Content, Pronunciation and Oral Fluency. Pronunciation and Oral Fluency are each rated 0–5 (Pearson Test Taker Score Guide).
- Pearson says humans never check pronunciation or fluency. Only the AI scores them.
- We found no microphone position or volume rule published by Pearson. Mic advice below is what test takers report, plus general audio sense.
Why is my PTE speaking score stuck at 53, 64 or 70 when I practise every day?
A plateau usually means one trait is capping you while the others improve. Speaking is scored on Content, Pronunciation and Oral Fluency, and three tasks also draw on your listening. Record your answers, find the lowest trait, and practise that one on purpose. More of the same practice rarely moves a score that one weak trait holds down.
- Content is counted word by word. On Read Aloud, each replaced, omitted or inserted word is one error.
- Repeat Sentence, Retell Lecture and Summarize Group Discussion count towards both Listening and Speaking, so a listening gap can look like a speaking plateau.
- Record every answer, listen back, and note what cost you: a sound, your pace or missing words. Fix one trait per week.
- A test taker on Reddit wrote that whichever sub-score is lowest across several tests is your ceiling. See how PTE is scored.
"Sir please tell me the reason of low marks in speaking i always got 53" (a YouTube commenter)
Why is my pronunciation score 10 when my oral fluency is 70+?
A very low Pronunciation score beside a high Oral Fluency score usually means your rhythm is fine but the engine could not match your sounds to what it expects. Quora answers blame microphone position, speaking too fast, flat pitch and background noise. Pearson has not confirmed any of these, so test one change at a time.
- Oral Fluency is flow: pauses, false starts, repetitions and rhythm.
- Pronunciation is, in plain terms, how close each sound is to a pattern the system recognises. Pearson says it asks whether you are understood, not whether you match one accent.
- Pearson says humans never check either trait. Nobody corrects for a poor recording.
- A test taker on Reddit wrote that the ending sounds, such as t and ed, need to be clearer.
- Record one sentence three ways, slower, with more varied pitch, and with the microphone moved, then compare the playback.
On supported speaking tasks, ScoreFluent shows four axes: Accuracy, Fluency, Prosody and Content. Tap a word to see a score for every sound in it, and replay that word from your own recording. For one common trap, see why your "th" costs pronunciation marks.
Where should the microphone sit and how loud should I speak?
We found no microphone position or volume rule published by Pearson, so treat all advice here as test-taker reports. Keep the microphone at a steady distance, speak at a steady level, and never speak so quietly that your recording is faint. Record yourself with the setup you will use and listen back before you trust it.
- One taker says a microphone below the mouth helped. Another says moving it changed nothing. Keep one steady distance.
- A YouTube commenter says never to reduce the microphone volume, only to increase it.
- A test taker on Reddit wrote that you should not breathe into the microphone: keep it near your chin and move it out a few centimetres.
Does a noisy test room lower my speaking score?
Pearson has not said how background voices affect scoring, so nobody outside Pearson can promise either way. One test taker describes a room of 15 people all speaking at once, and Quora answers disagree on whether the engine filters ambient noise. Practise with some noise, and keep your own volume and pace steady.
- Practise now and then with quiet background sound, such as a radio at low volume. A test taker on Reddit gave the same advice after a loud first exam.
"room of 15 people everyone shouting" (a Quora answer about a low speaking score)
Why do third-party apps give me 0 for pronunciation when I speak clearly?
Third-party apps use their own scoring engines, not Pearson's, so a 0 for clear speech usually means the app failed to match your voice, not that you spoke badly. Reviewers of competing apps report accent bias, correct words marked red and unstable scores. Compare the same recording across attempts and treat one score as a signal to check.
Reviewers of competing apps on Google Play and the App Store report:
- Accent bias: about a dozen reviewers of one app say UK, Indian, Pakistani and Persian speakers are marked low or not recognised. A test taker on Reddit wrote that they kept getting 10 out of 90 for fluency on one app, despite being obviously fluent.
- Read Aloud at 70 or more on a phone and about 65 on a laptop (two reviewers).
"Native English speaker from the UK, yet it rates me as only 47% fluent" (a reviewer of a competing app)
Record the same answer on two devices and compare. Treat any AI scorer as a guide, ours included. ScoreFluent scores are practice estimates, and we do not claim they match Pearson's.
Will the AI flag my Read Aloud or Describe Image template?
Pearson treats memorised material as irrelevant on Describe Image, and its score guide says heavily pre-prepared material that does not deal with the prompt can score zero for Content. Read Aloud has no template to flag, because the text is given. A short frame for your opening and closing is lower risk than a full script.
Pearson does not publish how it detects templates, so we cannot say what the AI flags. Reddit threads claim a human check follows any response flagged as a pure template. That is an unverified community claim. Keep any frame to one opening and one closing line, and fill the middle from the image or audio.
"using a generic template is dangerous as AI may flag it, flex template is harder than you think" (a YouTube commenter)
For the longer answer, read whether PTE templates still work in 2026.
Why won't the words come out in Repeat Sentence and Write from Dictation?
Both tasks play the audio once, so words vanish when you try to hold sounds instead of meaning. Chunk the sentence into phrases, repeat what you understood, and keep going. On Repeat Sentence, Pearson ignores hesitations and pauses when it scores Content, but they still cost you Oral Fluency.
- Repeat Sentence is 3–9 seconds of audio and 15 seconds to answer. Content is 3 only if every word is in sequence.
- Write from Dictation gives 1 point per correct, correctly spelled word, and counts towards Listening and Writing.
- Listen in chunks of three to five words. If exact recall fails after about 3 seconds, say the sentence you understood, smoothly. That is our technique, not a Pearson rule.
"I totally understand the question but in the time of answering my words will not coming" (a YouTube commenter)
More in our Repeat Sentence guide.
Does stammering or an accent lower my PTE score?
Pearson says PTE recognises regional and national varieties of English pronunciation, to the degree that most regular speakers can understand them. Pronunciation is judged on being understood, not on matching one accent. Hesitations, repetitions and false starts do count against Oral Fluency. Pearson's pages do not address stammering, so ask Pearson about test-day adjustments before you book.
- Accent bias reports about third-party apps concern those apps' engines, not Pearson's.
- A stammer can affect Oral Fluency, because its descriptors count hesitations and repetitions. We found no Pearson statement on stammering.
Which speaking tasks matter most for the speaking score?
By item count, Repeat Sentence is the biggest speaking task: Pearson lists 10 to 12 items, and it counts towards both Listening and Speaking. Read Aloud has 6 to 7 items and Describe Image 5 to 6. Pearson's score guide gives no per-task weight that we could find, so this shows where the most points are available.
| Task | Items per test | Counts towards |
|---|---|---|
| Repeat Sentence | 10–12 | Listening and Speaking |
| Read Aloud | 6–7 | Speaking |
| Describe Image | 5–6 | Speaking |
| Retell Lecture, Summarize Group Discussion | 2–3 each | Listening and Speaking |
| Respond to a Situation | 2–3 | Speaking |
Item counts are Pearson's published ranges, as listed on our PTE Speaking page. Answer Short Question is spoken but counts to Listening only. Use your own recordings to choose: a task with many items and a weak trait is the best place to start.
Practise this on ScoreFluent
ScoreFluent is an AI English coach for seven exam families. It turns each answer into clear feedback and the next thing to practise. For PTE Academic, all seven speaking task types are available. Answer Short Question, Read Aloud and Repeat Sentence are scored free within a daily limit. The other four get Personalised Tips, which cost 8 credits each, and new accounts start with 40 free credits. Start PTE practice.
This page was written by the ScoreFluent team from questions students asked in public. ScoreFluent is an independent practice platform and is not affiliated with, endorsed by or connected to Pearson. Practice scores are estimates, not official results.
Questions students ask
Why is my PTE speaking score stuck at 53, 64 or 70 when I practise every day?
A plateau usually means one trait is capping you while the others improve. Speaking is scored on Content, Pronunciation and Oral Fluency, and three tasks also draw on your listening. Record your answers, find the lowest trait, and practise that one on purpose. More of the same practice rarely moves a score that one weak trait holds down.
Why is my pronunciation score 10 when my oral fluency is 70+?
A very low Pronunciation score beside a high Oral Fluency score usually means your rhythm is fine but the engine could not match your sounds to what it expects. Quora answers blame microphone position, speaking too fast, flat pitch and background noise. Pearson has not confirmed any of these, so test one change at a time.
Where should the microphone sit and how loud should I speak?
We found no microphone position or volume rule published by Pearson, so treat all advice here as test-taker reports. Keep the microphone at a steady distance, speak at a steady level, and never speak so quietly that your recording is faint. Record yourself with the setup you will use and listen back before you trust it.
Does a noisy test room lower my speaking score?
Pearson has not said how background voices affect scoring, so nobody outside Pearson can promise either way. One test taker describes a room of 15 people all speaking at once, and Quora answers disagree on whether the engine filters ambient noise. Practise with some noise, and keep your own volume and pace steady.
Why do third-party apps give me 0 for pronunciation when I speak clearly?
Third-party apps use their own scoring engines, not Pearson's, so a 0 for clear speech usually means the app failed to match your voice, not that you spoke badly. Reviewers of competing apps report accent bias, correct words marked red and unstable scores. Compare the same recording across attempts and treat one score as a signal to check.
Will the AI flag my Read Aloud or Describe Image template?
Pearson treats memorised material as irrelevant on Describe Image, and its score guide says heavily pre-prepared material that does not deal with the prompt can score zero for Content. Read Aloud has no template to flag, because the text is given. A short frame for your opening and closing is lower risk than a full script.
Why won't the words come out in Repeat Sentence and Write from Dictation?
Both tasks play the audio once, so words vanish when you try to hold sounds instead of meaning. Chunk the sentence into phrases, repeat what you understood, and keep going. On Repeat Sentence, Pearson ignores hesitations and pauses when it scores Content, but they still cost you Oral Fluency.
Does stammering or an accent lower my PTE score?
Pearson says PTE recognises regional and national varieties of English pronunciation, to the degree that most regular speakers can understand them. Pronunciation is judged on being understood, not on matching one accent. Hesitations, repetitions and false starts do count against Oral Fluency. Pearson's pages do not address stammering, so ask Pearson about test-day adjustments before you book.
Which speaking tasks matter most for the speaking score?
By item count, Repeat Sentence is the biggest speaking task: Pearson lists 10 to 12 items, and it counts towards both Listening and Speaking. Read Aloud has 6 to 7 items and Describe Image 5 to 6. Pearson's score guide gives no per-task weight that we could find, so this shows where the most points are available.