PTE Speaking Retell Lecture 2026: Notes and Template

🦈 Retell Lecture is 40 seconds of unbroken speech under a three-second silence timer — the exact skill that daily conversation builds and silent study does not. SpeakShark gives you real spoken practice with phoneme-level pronunciation feedback while you talk, plus an AI teacher that remembers the mistakes you keep repeating between sessions. Free tier, no card. Start a free speaking session →

PTE Retell Lecture gives you a lecture of up to 90 seconds, 10 seconds to prepare, and 40 seconds to speak — and it is scored on three traits: Content out of 6, Oral Fluency out of 5, and Pronunciation out of 5, per Pearson's own Test Taker Score Guide. Because it is an integrated item, those points feed both your Speaking score and your Listening score, which makes it one of the highest-leverage tasks in the whole exam. This guide gives you the note-taking grid, the template that survives Pearson's anti-memorisation rules, and three original practice lectures — with every timing and rubric line traced to a Pearson document we opened, not to a coaching forum.

In this guide: at a glance · what Pearson actually scores · the note taking method · the template · filling the full 40 seconds · three original practice lectures · common mistakes · when this is the wrong prep · how we researched this · FAQ

PTE Retell Lecture at a glance

Two numbers in that table decide almost everything. Content is out of 6, not 5 — it is the single heaviest trait in the task, and it is the one a human checks. And 40 seconds is a floor as much as a ceiling: the fastest way to lose content points is to run out of things to say at second 22.

If you want the whole exam mapped before you drill one task, start with our PTE Academic speaking overview, then come back here.

What Pearson actually scores in Retell Lecture

Pearson's Test Taker Score Guide publishes the rubric in full. Here is what each band actually rewards, condensed from the official descriptors.

Content (0–6) — the heaviest trait

Read band 3 twice. It is the trap almost every self-taught candidate falls into: you heard the words, you wrote the words down, you said the words back — and the rubric calls that repeating language from the lecture without reformulation or context. Retell Lecture is a paraphrasing test wearing a listening test's clothes.

Oral Fluency (0–5) — measured in hesitations

Pearson's fluency descriptors are unusually literal, which makes them trainable:

• 5 Highly proficient — smooth rhythm and phrasing, no hesitations, repetitions, false starts or phonological simplifications
• 4 Advanced — acceptable rhythm and word emphasis, no more than one hesitation, repetition or false start
• 3 Good — acceptable speed, may be uneven, more than one hesitation but most words in continuous phrases, no long pauses, not staccato
• 2 Intermediate — uneven or staccato; at least one smooth three-word run; two or three hesitations; one long pause, but not two
• 1 Limited — irregular phrasing, poor rhythm, multiple hesitations and false starts
• 0 Disfluent — slow and laboured, most words isolated, more than one long pause

The gap between a 4 and a 2 is roughly two extra stumbles across 40 seconds. That is not a knowledge gap. That is a reps gap.

Pronunciation (0–5) — intelligibility, not accent

Pearson's own wording is about being understood, not about sounding American. A 5 requires vowels and consonants easily understood by regular speakers, with appropriate assimilation and deletion in continuous speech and correct word stress. A 2 means "at least 2/3 of speech is intelligible, but listeners might need to adjust to the accent." A 1 means listeners struggle with about a third of your words. A 0 means more than half is unintelligible.

So the target is not accent erasure — it is closing the specific consonant and vowel habits that cost you that third. That is exactly what phoneme-level feedback is for: not "practise pronunciation," but "your /θ/ is landing as /t/ in three words per minute." Hear which sounds are costing you — free session →

The scoring detail candidates keep getting wrong

Third-party coaching sites still circulate a "Content 0–5" figure for Retell Lecture. Pearson's own November 2024 score guide — the version we opened in July 2026 — publishes a six-band Content descriptor set, 6 down to 0. Where a prep site and the test maker disagree, use the test maker.

The second change is who reads your answer. The score guide lists Retell Lecture among the question types that "also include a human expert review of the Content before the final score for the task is finalized." Third-party trackers such as PTE Panacea date this change to 4 November 2024; Pearson's guide confirms the practice without dating it. The practical consequence is identical either way: a human decides whether your content is genuinely yours.

The note taking method: a three column grid

You are given an erasable noteboard booklet and pens at the test centre — Pearson's own test-day guide describes the booth as "a computer, keyboard, audio headset and an erasable booklet with two pens." Nothing else. So the method has to be fast, symbolic, and readable at a glance ten seconds later.

Pearson's Retell Lecture guidance tells you to "jot down key words and phrases" using punctuation like dashes, equals signs, ampersands, plus signs and arrows, plus bullet points. Here is that advice turned into a physical layout you draw before the audio starts.

Draw two vertical lines. Three columns, one strip across the bottom.

Three rules make it work:

• One line per idea, maximum four words. If a line takes longer than two seconds to write, you stopped listening to write it. Listening is the scored skill here too.
• Never write full sentences. A written sentence pulls you into reading aloud, and reading your notes verbatim reads as "repeating language from the lecture without reformulation" — the band-3 descriptor.
• Column 3 is what you actually get paid for. Pearson explicitly asks for "characters, actions, aspects, and relationships between the information." Columns 1 and 2 are the characters and actions. Column 3 is the relationships, and it is where band 5 and 6 answers separate from band 3 ones.

There is one more free second you are probably wasting. Pearson's official test tips note that you get three seconds to look at the image on screen before the recording begins — enough to predict the topic and pre-load vocabulary. Their worked example: a rocket image → rocket, launch, sky → topic space travel, exploration. Write those three words in the EXACT strip before a single word of audio plays.

The PTE Retell Lecture template that still works in 2026

First, the honest warning, because it now changes the answer to "what template should I use?"

Pearson states that if there is "a significant amount of pre-memorized content, a score of 0 will be given, and fluency and pronunciation will not be scored." A human expert checks Content specifically to catch this. So the era of the 15-word memorised opener that everyone in your coaching centre recites is over — not because it is banned outright, but because it burns 4 of your 40 seconds carrying zero lecture-specific content while flagging you to the exact reviewer looking for templates.

What survives is a sentence shape, not a sentence. A shape tells you what kind of clause comes next; you fill it with words the lecture just gave you. Here is the four-move frame, with a time budget.

Move 1 — Frame (about 6 seconds). Name the topic and the angle in one sentence. The angle is what makes it yours.

"The lecturer looked at [topic], and argued that [the lecturer's actual position]."

Move 2 — Three points, chained (about 20 seconds). One clause per point, joined by connectives. Pearson's guidance specifically asks for linking words such as so and then and relative clauses using which, who, that — those relative clauses are the cheapest way to show relationships without spending time.

"She began with [point 1], which meant that [consequence]. She then explained [point 2] — so [link]. The third part covered [point 3]."

Move 3 — One exact detail (about 8 seconds). Pull one number, name or example straight from the EXACT strip. This is the single highest-yield sentence in the answer, because it proves comprehension rather than recognition.

"For example, she mentioned [exact figure or name], which showed [why it mattered]."

Move 4 — Close on implication (about 6 seconds). Pearson's band-6 descriptor rewards "expanding on important points." A conclusion or implication is that expansion.

"Overall, the point was that [conclusion] — which suggests [implication]."

Notice that every bracket must be filled from your notes. The words outside the brackets total roughly 20 — small enough to be normal discourse, not a memorised block. If your template contains a whole clause that would make sense before you heard the lecture, cut it.

For the sister task with the same 40-second window and the same anti-template problem, see our PTE Describe Image guide — the structure logic transfers almost exactly.

How to keep speaking for the full 40 seconds

This is the mechanical problem, and it has a mechanical fix.

Pearson's official test tips are blunt about the trade-off: "You only have 40 seconds to speak and make sure that you include all the main points of the lecture. If you repeat ideas, correct yourself or hesitate, you will use up valuable time and lose score points. So keep talking and ignore any mistakes that you make."

Layered on top is the hard cut-off. Pearson's tips state that if you wait more than three seconds the recording stops and the item status changes to "completed," and that if you stop for three seconds while speaking, the status changes to completed "even if you haven't finished, and you won't be able to continue answering." The tips confirm this applies to Read Aloud, Repeat Sentence, Describe Image and Re-tell Lecture alike.

So three seconds of thinking silence ends your answer. Here is how to never spend it:

• Start on the tone, not on the thought. Your Move 1 frame is decided before the audio ends — you only need the topic word. Beginning within a second of the tone also protects you from the opening cut-off.
• Buy time with content, not with "um." Pearson's don'ts include filler words unrelated to the lecture. But a content-bearing bridge is not filler: "the second part of the lecture was more interesting, because…" moves the answer forward while your eyes find the next note line.
• Never self-correct. A repair costs two seconds and a fluency point, and the rubric counts false starts explicitly. Wrong word, wrong tense, wrong number — keep going. The AI is counting your hesitations, not grading your grammar in this trait.
• Chain with relative clauses. which, who, that let you extend a sentence indefinitely without a pause. A speaker who can add "…which meant that…" on reflex never falls into a three-second gap.
• Rehearse the last five seconds. Most candidates die at second 30 with nothing left. Your Move 4 implication sentence exists precisely to be deployed whenever you sense the notes running out — at second 30 or second 38.
• If you truly run dry, restate the relationship. "So the overall connection there was between [A] and [B]" is on-topic, uses a connective, and costs nothing.

The skill under all six bullets is the same one: producing continuous English while your attention is on the content, not on the language. You cannot build it by reading strategy posts, and you cannot build it silently. It is built by talking daily until the connectives come out on reflex — which is the whole design of SpeakShark's Daily Talk mode, where a 5-minute free session is roughly seven Retell-Lecture-lengths of unbroken speaking. And because the teacher carries your recurring mistakes between sessions, the day-10 session already knows which hesitation habit day-3 exposed. Build the 40-second reflex free →

If nerves rather than language are what freezes you at second 12, the graded-exposure method in our guide to overcoming the fear of speaking English is the better starting point.

Three original practice lectures you can run tonight

These are original practice materials written by us for speaking rehearsal. They are not real PTE items, they are not reproduced from any exam, and no real PTE content appears anywhere on this page. Each is roughly the length and density of a 90-second academic lecture, given as an outline so you — or a study partner, or an AI teacher — can speak it aloud as the prompt.

Practice lecture 1 — Why some cities are painting their roofs white

Outline to read aloud as the prompt: Dark roofs absorb sunlight and re-radiate it as heat, which is one reason dense city centres run warmer than the countryside around them. A coating of reflective white paint bounces a large share of that light straight back. Trials in several hot-climate cities have reported cooler indoor temperatures in top-floor flats and lower demand on air conditioning. But the effect varies: it is strongest on flat roofs in sunny, low-rainfall places, weaker where roofs are pitched, shaded, or quickly dirtied. Critics argue the money buys more cooling if spent on street trees, which also handle rainfall. The lecturer's position: white roofs are a cheap partial fix, not a substitute for shade.

Notes grid target: WHO/WHAT — dark roofs, white paint, trials, critics, trees. DOES/IS — absorb, reflect, cooler indoors, varies, costs. LINK — dark → heat x · paint → cooler · flat + sunny = best · trees ≠ roofs · ∴ partial fix.

Model 40-second shape: Frame the topic and the lecturer's "cheap partial fix" position → three points (why dark roofs heat cities, what the paint does, where it fails) → one exact detail (the flat-roof-and-low-rainfall condition) → close on the implication that shade planting is the complement, not the rival.

Practice lecture 2 — The problem with measuring productivity

Outline: Productivity is normally defined as output per hour worked, which is straightforward in a factory and slippery everywhere else. When the output is a diagnosis, a piece of software or a lesson, the numerator is hard to count and quality changes are invisible in the statistics. Economists have long noted that measured productivity growth slowed in advanced economies even as digital tools spread — the so-called paradox. Two explanations compete: either the tools genuinely deliver less than promised, or the gains are real but land in things national accounts do not capture, such as convenience and free services. The lecturer's position: measurement error explains part of the gap, but not all of it.

Notes grid target: WHO/WHAT — productivity, factories, services, economists, digital tools. DOES/IS — output ÷ hours, easy vs hard to count, growth slowed. LINK — services → quality invisible · tools + slow growth = paradox · explanation 1 ≠ explanation 2 · ∴ partly measurement.

Model shape: Frame the definition and why it breaks → the two competing explanations as your chained points → one exact detail (the paradox, digital tools versus slowing growth) → close on the lecturer's "part but not all" verdict.

Practice lecture 3 — What the brain does while you sleep

Outline: For most of the twentieth century sleep was treated as the brain switching off. Research over the last two decades has reframed it as an active maintenance shift. During deep sleep, the spaces between brain cells appear to widen, and fluid moves through the tissue more freely, carrying away waste products that build up during waking hours. Separately, patterns of electrical activity recorded during sleep resemble replays of the day's experiences, which is one proposed mechanism for how short-term memories become durable ones. Both findings are still debated in detail. The lecturer's position: the switching-off metaphor is dead, but the replacement story is not finished.

Notes grid target: WHO/WHAT — old view, deep sleep, fluid, replay, memory. DOES/IS — off vs active, spaces widen, clears waste, repeats day. LINK — old view x · deep sleep → clearance · replay → memory durable · both = debated · ∴ story unfinished.

Model shape: Frame the old-view-versus-new-view angle → clearance and replay as two chained points → one exact detail (spaces widening during deep sleep) → close on the "unfinished" implication.

Run each one three times across a week, not three times in an hour. Rep one exposes what you miss; rep two fixes the notes; rep three is where fluency actually moves. The same rotation logic applies to the other spoken tasks — see PTE Read Aloud and PTE Repeat Sentence, which train the two sub-skills Retell Lecture depends on: clean continuous delivery and accurate short-term auditory capture.

Common Retell Lecture mistakes and what they cost you

If your score report already came back low and you cannot tell which of these caused it, our diagnostic walkthrough in why your PTE speaking score is low maps symptoms to traits.

When Retell Lecture practice is the wrong thing to do

An honest guide has to say where this advice stops.

If your listening is the bottleneck, drilling the template is wasted effort. You cannot retell a lecture you did not follow. If you routinely reach the 10-second prep window with three shaky note lines, your problem is comprehension speed, not structure. Fix that with volume — podcasts, lectures, unscripted conversation at natural pace — before touching templates. Our free English speaking level test gives you a rough baseline first.

If one sound is your problem, a drill app beats a conversation app. Tools built for isolated phoneme repetition with mouth-position graphics do a job that no conversational tool, ours included, does as sharply. Use one for a stubborn /v/–/w/ or /θ/–/t/ habit, then bring it back into continuous speech.

If you need the exam format itself, go to Pearson. We are not an exam-prep provider and we do not sell practice tests. Official scored practice, real item formats and the current score guide come from Pearson directly. What we can build is the underlying spoken fluency that the rubric measures — which is the part official practice tests diagnose but do not train.

If your goal is a job rather than a visa, you may be preparing the wrong test. Plenty of candidates drill PTE when the actual gate is an employer's automated spoken-English assessment. If that is you, our guides to the Versant test at Accenture and how to prepare for a Versant test cover a different format with a familiar engine — Pearson's own glossary describes Versant technology as the "proprietary speech processing system" designed to analyse and automatically score speech. Same lineage, different exam.

And if the retell itself is fine but short answers collapse, the constraint is different again — see PTE Answer Short Question, where brevity, not endurance, is the scored skill.

How we researched this guide

Every timing, band and quotation above comes from a page or PDF we opened in July 2026, linked so you can verify it.

Primary sources — Pearson, the test maker. The PTE Academic and UKVI Speaking & Writing test format page for the 10-second preparation, 40-second response, up-to-90-second lecture, the Listening-and-Speaking skill attribution, and the 76–84 minute length of Part 1. The PTE Academic Test Taker Score Guide (November 2024, v4) for the 2–3 item count, the partial-credit scoring type, the full Content 0–6 descriptors, the Oral Fluency and Pronunciation 0–5 descriptors, the 10–90 score range, the human review of Content, and the anomaly-scoring rule for quiet or unclear responses. Pearson's article on answering Retell Lecture effectively for the note-taking symbols, the "characters, actions, aspects, and relationships" requirement, the linking-word and relative-clause advice, and the pre-memorised-content penalty. Pearson's official PTE Academic Test Tips resource for the 3-second image preview, the "keep talking and ignore any mistakes" instruction, and the 3-second silence cut-off that applies across the speaking items. Pearson's test-day guide for the booth setup and the erasable booklet with two pens.

Secondary source, treated as a claim. PTE Panacea and similar coaching sites date the introduction of human Content review to 4 November 2024. Pearson's score guide confirms the practice but does not print that date, so we attribute the date to the third-party sources rather than to Pearson.

What we deliberately did not do. We did not reproduce a single real PTE item. The three practice lectures on this page are original, written from scratch for speaking rehearsal, and the templates are sentence shapes rather than scripts — partly on principle, partly because Pearson's stated penalty for memorised content makes recycled scripts actively dangerous. We also did not publish a "score X for a 79+" table: Pearson publishes trait bands, not per-item targets, and inventing one would be guesswork dressed as data.

SpeakShark is a speaking-improvement tool, not an exam-prep provider, and is not affiliated with or endorsed by Pearson, PTE, IELTS, OET, Cambridge Boxhill, HireVue, or any employer or institution named on this page.

FAQ

Retell Lecture rewards one thing above all others: the ability to keep producing organised English for 40 uninterrupted seconds while your attention is somewhere else entirely — on the notes, on the meaning, on the clock. That is a reflex, and reflexes are built by speaking every day, not by reading rubrics.

Start your free daily speaking session on SpeakShark → — three 5-minute sessions a day with Sarah across 64 topics, no card, with pronunciation feedback while you speak and a teacher that remembers what you got wrong yesterday. Premium is $10/month or $69/year (about $5.75/month) for 10-minute sessions with unlimited turns, four teachers in US, UK, Australian and Canadian accents, and 320 topics — details on pricing.

All Pearson timings, item counts and rubric bands verified July 2026 from the official documents linked above. Practice lectures on this page are original material written for speaking rehearsal, not exam content. Test formats and scoring policies change — confirm current details on Pearson's own site before your test date.