TOEFL Speaking Section Guide

TOEFL Speaking is the last section on test day, after Reading, Listening and Writing. It is linear: every test taker answers the same 11 items in about 8 minutes. The current format took effect worldwide on January 21, 2026, when ETS replaced the older Speaking tasks and moved the whole exam onto a new 1 to 6 score scale.

Two task types make up the section: Listen and Repeat and Take an Interview. Software scores both tasks, not a human rater, and the result feeds into a Speaking score on a 1 to 6 scale in half-point steps. This guide walks through the section structure, both task formats, how to use your response time, and how the scoring works. It closes with where to find live Speaking practice on 10exam.

Speaking Section Structure and Timing

Three numbers describe the section.

ETS sets the Speaking section at 11 total items in about 8 minutes, split into 7 items for Listen and Repeat and 4 items for Take an Interview. Reading and Listening adapt to your answers as you go. Speaking does not: every test taker gets the identical set of tasks in the identical order.

Speaking comes last in a fixed section order: Reading (50 items, about 30 minutes), Listening (47 items, about 29 minutes), Writing (12 items, 23 minutes), then Speaking, with no scheduled break between sections. You reach the microphone at the end of a 1 hour 23 minute to 1 hour 29 minute test, down from close to 2 hours before the redesign, so plan your energy for a late finish. For the full breakdown of every section, see our TOEFL iBT exam guide.

Task 1: Listen and Repeat

Listen and Repeat is the first Speaking task, with 7 items. You hear a spoken sentence inside an academic or campus-life scenario, then repeat it back exactly as you heard it. The task measures intelligible, accurate reproduction of spoken English, and its target range spans the full CEFR scale, from A1 to C2.

Sentences start short and simple with a single independent clause, then progress to longer utterances with compound verbs, dependent clauses and relative clauses. Early items in the set are easier to reproduce word for word. Later items pack in more clauses, so accurate word order and word endings carry more weight.

How to Approach Listen and Repeat

A few habits protect your score on this task.

• Listen for the whole sentence before you speak. Starting mid-sentence breaks your word order and costs accuracy.
• Hold onto sentence structure, not only isolated words. Subject, verb and clause order carry weight on their own.
• Match the stress and pausing of the original sentence. A natural rhythm scores better than a flat, word-by-word recitation.
• Keep small words intact. Articles, prepositions and verb endings are easy to drop under pressure. Dropping them lowers the accuracy and intelligibility ratings behind your band, even when the basic meaning survives.

Task 2: Take an Interview

Take an Interview is the second Speaking task, with 4 items. You hold a simulated interview with a prerecorded interviewer, tied to academic or campus situations. Like Listen and Repeat, the CEFR target spans A1 to C2.

The 4 questions build in difficulty. Early questions ask for brief factual statements, and later questions ask for more elaborate expressions of opinion, with difficulty rising across the task. ETS does not publish the interview scenarios, so use the shape rather than the topic. As an illustrative example of shape only: an interviewer might first ask which subject you study, then ask which part of it interests you most and why.

How to Approach Take an Interview

Score-carrying habits differ from Listen and Repeat, since this task is unscripted.

• Answer the question asked, not the topic in general. A specific answer to a specific question scores higher than a rehearsed speech on the scenario topic.
• Give a reason with every opinion. Later questions reward elaboration, so a bare opinion without support leaves points unclaimed.
• Use full sentences, not fragments. The interview format rewards connected, natural speech over a list of keywords.
• Stay inside the scenario. Each question sits in the same academic or campus situation, so keep your answers consistent with what you said earlier.

Response Time and Delivery

ETS publishes 11 items in about 8 minutes for the section and no per-item response window. Our own arithmetic on those two published numbers leaves an average under 45 seconds per item, and the average has to absorb the audio you hear, your answer and the gap between items. Every Speaking response is short by design.

Test-prep publishers outside ETS quote tighter windows. ETS has not confirmed them on an ETS-hosted page or in the 2026 blueprint, so treat the row values below as a planning estimate rather than a rule.

The same non-ETS sources report zero preparation time before either task, which again is unconfirmed by ETS. For the on-screen timing you will face, work from the official ETS Speaking page, which carries the current task descriptions and the sample material ETS makes available.

Delivery Strategy

Short windows reward a controlled start and a steady finish. The four habits below are our coaching advice, not ETS scoring rules.

• Start speaking as soon as the response window opens. Silence at the front of a short window burns the seconds you need at the end. Speaking reports one band built from item scores, with no separate fluency mark, so the cost of a slow start is lost words rather than a lost sub-score.
• Speak at a steady, natural pace. Rushing to fit more words in produces run-on speech, which is harder to score as accurate.
• Plan your first sentence while you listen. For Take an Interview, form your opening line before the response window starts, then build the rest as you speak.
• Use the whole window. The blueprint expects elaboration on familiar topics at B1 to B2 and fluent, spontaneous answers at C1 to C2, so a developed answer with a reason beats a one-line reply. Stop once your point is complete rather than repeating yourself.

How Speaking Is Scored

Both Speaking tasks are scored by software using natural language processing and machine learning, not by a human rater. Each item earns up to 5 raw points, for a maximum of 55 raw points across the section: 7 items worth 5 points plus 4 items worth 5 points.

ETS converts your raw score into a 1 to 6 Speaking score in 0.5-point steps. One band covers the whole section, so there are no published sub-scores for pronunciation, fluency or grammar. The scale is CEFR-aligned end to end: a 6 lines up with CEFR C2 and a 1 lines up with CEFR A1. This replaces the old 0 to 30 Speaking scale.

The 2026 blueprint sets out what each CEFR band sounds like in Speaking.

These bands come from the 2026 blueprint's Speaking claim table. Use them to judge where your current speech lands. A response landing at B1 to B2 shows mostly accurate pronunciation and elaboration on familiar topics. A response landing at C1 to C2 shows fluent, spontaneous replies to a harder question, with nuanced meaning.

Your Speaking score is one of four section scores. Your overall TOEFL score is the average of your Reading, Listening, Speaking and Writing scores, rounded to the nearest half band. A weak Speaking score pulls the overall average down by the same weight as any other section. For how Speaking difficulty stacks up against the rest of the test, see how hard the TOEFL iBT is.

Practice Speaking on 10exam

Full Reading and Writing practice is live on 10exam now, with real timing, real item counts and instant explanations. Every mock and drill is free, with no signup required. Speaking practice is on our roadmap and not yet available.

Use our TOEFL exam page to see what is live today, and start with Reading and Writing while Speaking content is built out.

Sources

• ETS: TOEFL iBT Speaking Section
• ETS: TOEFL iBT Test, 2026 Update Blueprint and Specifications Document
• ETS: TOEFL iBT Score Scale Update