TOEFL Listen and Repeat: Complete Guide | TOEFL 2026 New Format
Listen and Repeat is one of the new question types introduced in the TOEFL iBT® overhaul that took effect on January 21, 2026. The Speaking section was fully redesigned, and Listen and Repeat is now the first speaking task. It did not exist on the old test, so if you prepared before 2026 it will be new to you.
The task sounds easy: you hear a sentence and say it back. The challenge is that you hear it only once, with no text and no time to prepare. This guide explains exactly what Listen and Repeat is, how it is scored using the official ETS numbers, and the strategies that raise your score.
Table of Contents
- What is the "Listen and Repeat" question type?
- How "Listen and Repeat" is scored
- Tips to do well on "Listen and Repeat" questions
- How to practice "Listen and Repeat" questions
- Frequently asked questions
- Conclusion
What is the "Listen and Repeat" question type?
According to ETS, in Listen and Repeat you "listen to short sentences and repeat them exactly as you hear them." The task measures your ability to process spoken English and produce speech that is accurate and clearly intelligible. There is no reading, no opinion, and nothing to invent.
Listen and Repeat is the first of two tasks in the 2026 Speaking section, followed by Take an Interview. You hear seven separate sentences. Each one plays a single time with no text on screen, and after a beep you record yourself saying it back. The sentences use everyday academic and campus language, and they get longer as the task goes on.
Here is how a single item works:
After the beep, you say the same sentence back, clearly and accurately: "The library extends its opening hours during the final exam period."
For more worked examples like this one, see our 50 Listen and Repeat practice questions with answers, which cover the full range of contexts and patterns you may face.
How "Listen and Repeat" is scored
Listen and Repeat is scored differently from the typing tasks, so be clear on how points work. The official ETS 2026 Test Blueprint states:
- There are 7 Listen and Repeat sentences.
- Every response is scored by AI, not by a machine match or a human rater.
- The maximum is 5 points per sentence, so the task is worth up to 35 points.
- The target level runs across the full range, from A1 to C2.
The AI scores how accurately and clearly you reproduce the sentence, not whether you sound like a native speaker. You are not judged on your accent. A response that is clear and complete scores well even with a noticeable accent. A response that is fast but slurred, or that trails off into silence, does not.
Listen and Repeat sits in the Speaking section, which ETS estimates at about 8 minutes in total. ETS does not publish a separate time limit for each sentence. Some prep sites list exact per-sentence windows, but treat those as estimates, not official: in practice you get a short window of roughly 8 to 12 seconds to respond after the beep, and the later sentences allow a little more time because they are longer.
Tips to do well on "Listen and Repeat" questions
Listen to the whole sentence before you speak
You hear each sentence only one time, so give it your full attention from the first word. Do not start planning your answer while the audio is still playing, or you will miss the end. Wait for the beep, then begin.
Do not copy the speaker's speed or accent
The recording may be fast and smooth. You do not need to match it. The score is based on accuracy and clear speech, not on sounding like a native speaker. Speaking a little slower so every word is clear is better than rushing and slurring.
Break the sentence into chunks as you listen
Hold the sentence in memory by grouping it: who or what the sentence is about, the main action, then the extra details. Remembering three chunks is easier than remembering ten separate words.
Never go silent, even if you miss a word
If you forget part of the sentence, do not stop. Keep speaking and finish with the best version you can. A complete attempt with one wrong word scores far better than half a sentence followed by silence.
Repeat the exact words, do not paraphrase
The task asks for the sentence exactly as you heard it. Changing words or rephrasing lowers your accuracy score even if your version means the same thing. Reproduce the sentence, do not rewrite it.
Copy the stress and rhythm, not only the words
Clear speech includes which words are stressed and where the sentence rises and falls. Try to echo that pattern. It makes you easier to understand and supports the intelligibility part of the score.
Speak clearly into the microphone at a steady volume
Check your microphone during setup. Speak at a normal, steady volume, not too quiet and not so loud that the sound distorts. The AI can only score what it can clearly hear.
Expect the later sentences to be longer
The seven sentences increase in length and difficulty. Do not be surprised when the last few are harder. The response time is a little longer for them, so stay calm and use the chunking method.
If you forget a word, give a grammatical best attempt
When a word is gone, replace it with something that keeps the sentence grammatical and finish smoothly. A fluent, complete sentence with a small error sounds clearer than a broken one.
Train with hear once, pause, record, compare
In practice, play each sentence a single time, pause, and record yourself. Then listen back and compare your words, stress, and rhythm with the original. This builds the fast memory the task needs.
How to practice "Listen and Repeat" questions
Replaying the same easy sentences many times does not build the skill this task needs. Two things make the difference: practicing at your own level, and learning from the sentences you get wrong. Because you only hear each sentence once on test day, your practice has to train memory under the same condition.
Arno makes both of those easy, and free. You practice Listen and Repeat sentences matched to your level instead of a random set that is too easy or too hard, and you get feedback on what you got wrong so each miss teaches you something. That is how you turn the tips in this guide into real points on test day.
Click here to create your free account!
Frequently asked questions
How many Listen and Repeat sentences are on the TOEFL?
There are 7 sentences. Listen and Repeat is the first of two tasks in the 2026 Speaking section, followed by Take an Interview.
How is Listen and Repeat scored?
Each sentence is scored by AI on a 0 to 5 scale, for up to 35 points in total. The score reflects how accurately and clearly you reproduce the sentence.
Does my accent affect my Listen and Repeat score?
No. You are not scored on sounding like a native speaker. The score is based on accuracy and clear, intelligible speech, so a clear response with an accent scores well.
How many times do you hear each sentence, and how long do you get to answer?
You hear each sentence one time only, with no text on screen, and you respond after a beep. ETS does not publish official per-sentence times; the whole Speaking section is estimated at about 8 minutes.
What does Listen and Repeat measure?
Your ability to process spoken English and produce speech that is accurate and clearly intelligible. It does not test opinions, reading, or content knowledge.
Is Listen and Repeat new on the 2026 TOEFL?
Yes. It was introduced in the TOEFL iBT update that took effect on January 21, 2026, when the Speaking section was redesigned. It did not appear on earlier versions of the test.
Conclusion
Listen and Repeat rewards focused listening and clear speech. There are 7 sentences, AI scores each one out of 5, and you hear every sentence only once. Give the audio your full attention, repeat the exact words clearly without chasing a native accent, and never let a missed word turn into silence. With steady practice at the right level, this becomes one of the more predictable parts of the Speaking section.