TOEFL iBT Speaking Section 2026: Task Types, Scoring & Audio + Video Samples

What the Speaking section measures

Speaking measures both foundational language skills and meaningful communication. The ability to process language and produce fluent, intelligible speech is measured by a task where you repeat spoken input; more complex communication — expressing and supporting opinions and experiences — is measured by a simulated conversation. Like Writing, Speaking is not adaptive: everyone answers the same set of tasks. Every task involves listening first, and there's no preparation time, so responses are spontaneous.

Responses are scored holistically — one overall score per response — with attention to Delivery (clarity, pronunciation, pacing, intonation), Language Use (grammar and vocabulary), and Elaboration (how clearly and fully you support your ideas). Spontaneous speech isn't expected to be perfect; even high-scoring responses can contain occasional slips.

Task 1: Listen and Repeat

You take part in a realistic scenario — for example, being trained to help students request official transcripts — and hear seven sentences embedded in that context, repeating each one only once, immediately after hearing it. The sentences increase in complexity: two short ones first (about 9–11 syllables, 8 seconds to repeat), then three medium ones (14–16 syllables, 10 seconds), then two long ones (19–23 syllables, 12 seconds). Each sentence is paired with a visual cue that grounds the scenario. This task is scored 0–5 on two criteria: Intelligibility (how clearly you pronounce the words) and Accuracy (how closely your repetition matches the original). Partial credit is awarded for any part you repeat correctly.

Sentence: "Fill out the form to request your transcript." (a short, 10-syllable sentence — one clause, no modifiers.)

Sentence: "If you need an unofficial transcript, which you can keep for your own records, check this box." (a long, 22-syllable sentence with a conditional clause and a relative clause.)

Strategies for Listen and Repeat:

• Wait for the beep before speaking so your full response is recorded.
• Finish the sentence even if you slip. A mispronunciation isn't the end — keep going to the end, because partial credit is awarded for any part repeated accurately.
• Repeat only once — extra time at the end after you speak is fine.
• Use the visual cue to anchor the context, but don't let it distract you if it isn't helpful.
• Practice repeating sentences from an audiobook or online video; the better you understand simple and complex structures, the easier they are to recall and reproduce.

Task 2: Take an Interview

You take part in a simulated conversation with a virtual on-screen interviewer in a realistic context — applying for a scholarship, joining a research study. You answer four questions that progress in complexity: early questions cover personal experiences and factual information; later ones ask you to share opinions, speculate, or respond to hypotheticals. You get 45 seconds per response and no preparation time. It's scored 0–5 on four dimensions: Relevance (how well you address the question), Elaboration (how fully you support your ideas), Delivery (fluency, pacing, rhythm), and Language Use (grammar and vocabulary).

Question: "When you have a choice to make, do you usually decide quickly or take your time to think it over?"

What makes it a 5: the speaker clearly addresses the question, distinguishes quick choices for small matters from careful thought for important ones, and elaborates with specific examples (choosing a course vs. picking a snack). The pace is natural with authentic pauses and fillers, pronunciation is clear, and grammar and vocabulary are varied and accurate — minor slips typical of a C1-level speaker don't impede understanding.

The two families of interview questions

Personal questions ask about your experiences, plans, preferences, and people close to you — "Can you tell me about a time when your daily routine was disrupted?" or "What are your career goals after you graduate?" Abstract and speculative questions ask for opinions on current topics, predictions about the future, or answers to hypotheticals — "Do you think students learn better in groups or individually?" or "Imagine you're unclear about a concept in class — would you ask a classmate or your professor? Why?"

Strategies for Take an Interview:

• Use your own words, not memorized sentences — canned language scores lower. Answers that sound rehearsed rather than genuine are penalized.
• Address the question directly, then support your point with a specific reason or example.
• Speak at a natural pace with the full 45 seconds in mind; a little hesitation and self-correction is normal and doesn't hurt.
• Practice the common question types — describing experiences, stating preferences with reasons, giving opinions, and handling hypotheticals — so the format feels familiar under time pressure.

Complete Chapter 5 speaking audio lab

The samples above demonstrate Tracks 37, 42, 45, and 49. This lab adds every remaining Chapter 5 track so you can work through the complete instructional sequence without duplicating those players.

Listen and Repeat — full progression

Track 36 — scenario and instructions

You are helping students request transcripts. Listen for the scenario, then repeat each sentence once after the beep.

1. Track 38
   Check the target sentence

   Log into the portal and check your records.

2. Track 39
   Check the target sentence

   Ensure your name on the form matches the records before submitting.

3. Track 40
   Check the target sentence

   We will stamp your transcript with our seal before mailing it.

4. Track 41
   Check the target sentence

   Your transcript will be mailed directly to your selected schools.

5. Track 43
   Check the target sentence

   Email the registrar’s office to resolve issues, such as missing courses, right away.

Tracks 37 and 42 appear in the Listen and Repeat explanation above. Together, Tracks 37–43 progress from short independent clauses to longer sentences with modifiers, conditional clauses, and embedded phrases. Score each attempt for intelligibility and accuracy; repeat what you remember even when recall is incomplete because partial credit is possible.

Take an Interview — complete four-question sequence

Track 44 — scenario and instructions

The interview is a university study about everyday decision-making. Track 45 and its model response, Track 49, appear above. Continue with the remaining three questions and model responses:

1. Question — Track 46

   Describe a simple recent decision, such as choosing food or clothing, and explain the factors you considered.

   Model response — Track 50

   Why the model is effective

   The model weighs hunger, time, health, convenience, and cost before choosing an omelet. It stays relevant, develops specific support, maintains a conversational pace, and uses sufficiently varied language.

2. Question — Track 47

   Describe a time you asked someone for advice, the situation, and how the advice helped.

   Model response — Track 51

   Why the model is effective

   The model explains asking a technology-savvy classmate for phone advice and connects the recommendation to budget and needs. It stays relevant, develops specific support, maintains a conversational pace, and uses sufficiently varied language.

3. Question — Track 48

   Should people decide independently or seek advice from others? Explain your view.

   Model response — Track 52

   Why the model is effective

   The model takes a balanced position: seek advice for important decisions while practicing independence on smaller choices. It stays relevant, develops specific support, maintains a conversational pace, and uses sufficiently varied language.

Record-and-review checklist

• Did I address every part of the question directly?
• Did I add a reason, example, or consequence rather than repeating the prompt?
• Was my pace natural, with pauses that did not break comprehensibility?
• Were pronunciation, rhythm, and intonation easy to follow?
• Did grammar and vocabulary let me express precise meaning?

There is no preparation time and responses cannot be rerecorded. A natural accent is acceptable; pronunciation affects the score only when it reduces intelligibility. If you finish early, add a relevant detail rather than repeating yourself.

Full Practice Test — Speaking

11 questions across two task types. There is no preparation time; on the real test you speak into a microphone. Use a phone or recorder to practice, and time yourself. Reproduced from the official practice materials.

Listen and Repeat (1–7)

Instructions — Track 82

Scenario: you are volunteering in a student resource room; the manager is training you to help students use office equipment. Listen and repeat each sentence only once, after the beep.

1. (repeat once, ~8s)
Target sentence: "Print with various paper sizes."

2. (repeat once, ~8s)
Target sentence: "You can make numerous copies quickly."

3. (repeat once, ~10s)
Target sentence: "The scanner converts documents into digital files."

4. (repeat once, ~10s)
Target sentence: "The fax machine can send pages quickly over phone lines."

5. (repeat once, ~10s)
Target sentence: "The shredder securely destroys sensitive information."

6. (repeat once, ~12s)
Target sentence: "All devices are located in the office for easy use by those who need them."

7. (repeat once, ~12s)
Target sentence: "When in doubt, read the maintenance instructions that are available for all equipment."

Take an Interview (8–11)

Instructions — Track 90

Scenario: you volunteered for a research study about people's experiences with art and creativity. Watch each video, then answer the interviewer (45 seconds each, no prep time).

8.

9.

10.

11.

Score your responses against the 0–5 Listen-and-Repeat (Intelligibility + Accuracy) and Take-an-Interview (Relevance, Elaboration, Delivery, Language Use) rubrics in the Speaking guide above.