Listen and Repeat opens the Speaking section. You hear a short sentence once and say it back — every word, in the same order, with the same endings. It sounds simple until the sentences grow longer, and the words that disappear are almost always the small ones: an article, an -s, a before, the last phrase.
Each set begins with a one-line scene, such as a tutor showing new students around a piece of equipment. Then seven sentences about that scene play one at a time. You hear each one once, and the response window begins straight away. You never see the sentence on screen.
The early sentences are short single clauses. The later ones add places, times and linked clauses (before you…, until…, if…), or join two instructions with and. In our course we give short response windows of around 8 to 12 seconds depending on sentence length; that is a practice setting, so on test day follow the timer on screen.
The reference is the exact sentence. A near-synonym, a moved phrase or a dropped ending all change the sentence, even when the meaning survives. An accent is not a problem as long as each word is there and understandable.
Repeating a sentence accurately depends on hearing its structure, not just its sounds. Listen and Repeat checks:
Trying to hold fifteen separate words is close to impossible; holding three or four meaningful chunks is not. Spoken English arrives in thought groups — who or what, the action, where, when and why — and the connectors when, because, before, until and and usually start a new one.
As the sentence plays, notice the groups and count them as it ends. Then each group can bring back its own small words: if you remember on the floor as one unit, the comes back with it. Listening word by word keeps only the loud content words, and the rebuilt sentence sounds like a telegram.
Your memory of a sound is sharpest in the first second or two and fades quickly. Waiting to “get it right” only makes it worse, so begin the first group as soon as recording starts. Deliver each thought group in one smooth run, with only a brief break before the next, at an even, natural pace — slow, over-careful delivery gives the end of the sentence time to fade.
If a word is gone, keep the frame around it (in the … room) and carry on. A gap costs one word. Stopping to start again spends the time and memory you need for the last groups, which is where most words are lost anyway.
The scene line tells you the topic before any audio plays, so many content words become predictable. That frees your attention for the parts that really go missing: articles, prepositions, auxiliaries and endings.
Grammar is a second safety net. A singular subject in the present needs -s (The gym opens at six). Yesterday demands a past -ed, even if the sound was tiny. After let or watch plus an object, the next verb has no ending. Watch for the nearest-noun trap — in The chairs in the back room belong to…, the verb agrees with chairs, not room — and for hidden auxiliaries like the ’s in Somebody’s left the window open.
In the longest items, catch two things while listening: the main verb and the hinge word where the second half begins (before, until, if, so that, and, but). Together they split the sentence into two parts with clear jobs, and the smaller details hang on those parts.
Let the hinge predict what follows: before signals that an action is coming, until an end point, if a condition. Knowing what kind of words are due makes the end of the sentence far easier to recall. Finally, check the very last words — short endings such as again, first or by the door are the first to fall away.
English often lets a time phrase or a when-clause sit at either end of a sentence. Both orders are correct English, but only the one you heard is correct here. The same goes for near-synonyms: finish for ends, near for next to, collect for pick up. Your brain stores meaning faster than sound, so a swap feels right — keep the sound of each group’s key word instead.
Small modifiers carry meaning too. Only members is not members; the largest room is not the large room; put it back is not put it. Each one is a word the response must keep.
These errors account for most lost words in Listen and Repeat. Each comes with the fix and the reason it matters.
Mistake: Repeating “Printer on first floor needs more paper.”
Better: “The printer on the first floor needs more paper.”
Why: Two articles are gone. The idea survives, but the task checks the sentence, not the idea.
Mistake: Hearing “She locks the lab at six” and saying “She lock the lab at six.”
Better: Keep the -s: “She locks the lab at six.”
Why: An ending is part of the word. A dropped or added ending counts as a changed word.
Mistake: Hearing “When the music stops, take your seats” and saying “Take your seats when the music stops.”
Better: Keep the clause where you heard it.
Why: Both versions are good English, but they are different sentences. The order is part of the answer.
Mistake: Summarizing a long sentence: “People who borrow games must bring them back before midnight.”
Better: Rebuild the exact sentence group by group.
Why: A summary keeps the meaning but changes most of the words.
Mistake: Pausing after small words: “Please put your bag on the | floor by the | door.”
Better: “Please put your bag | on the floor | by the door.”
Why: A pause after the cuts an article off its noun. Pauses belong between thought groups.
Mistake: Stopping after a complete-sounding first half: “Return the classroom keys to the front desk.”
Better: Carry on through the hinge: “…before the evening class starts, and sign the sheet by the door.”
Why: In long sentences, more than half the words can come after the first main clause.
Two sets from our free practice bank, each trimmed to sentences of different lengths so you can see how the risks change as sentences grow. Play the audio first and try each sentence yourself before reading the sentence and the notes.
A media tutor introduces a simple stop-motion animation station.
Q1Listen, then repeat the sentence exactly.
Sentence: The model begins inside the square.
Two groups: “The model begins” / “inside the square”. Learners often drop the -s of “begins” or shorten “inside” to “in”. Stress falls on MODel, beGINS and SQUARE.
Q2Listen, then repeat the sentence exactly.
Sentence: Keep hands outside the camera view when each frame is captured.
Two groups: “Keep hands outside the camera view” / “when each frame is captured”. There is no “your” before “hands”, though learners often add one, and in the passive “is captured” the weak “is” and the final -d are easily lost. Stress falls on KEEP, HANDS, CAMera, VIEW, EACH, FRAME and CAPtured.
Q3Listen, then repeat the sentence exactly.
Sentence: Save the short practice sequence before clearing the table, and leave the camera in its marked position for the next group.
Two halves: “Save the short practice sequence before clearing the table” / “and leave the camera in its marked position for the next group”. The small modifiers “short” and “marked” are the words most often dropped, “its” tends to become “the”, and the t of “next” is barely said before “group”. Stress falls on SAVE, SHORT, SEquence, CLEARing, TAble, LEAVE, CAMera, MARKED, poSItion, NEXT and GROUP.
Audio in these samples uses AI-generated voices.
A media helper explains a practice upload interface.
Q1Listen, then repeat the sentence exactly.
Sentence: A completed transfer does not finish processing.
Two groups: “A completed transfer” / “does not finish processing”. Say “does not” in full, not “doesn't”, and keep the -ed syllable of “completed”. Stress falls on comPLETed, TRANSfer, NOT, FINish and PROcessing.
Q2Listen, then repeat the sentence exactly.
Sentence: If the preview fails, write down the job number before you try the upload again.
Three groups: “If the preview fails” / “write down the job number” / “before you try the upload again”. The -s of “fails” is easy to lose, and “the upload” is a noun here, stressed on UP-, not like the verb upLOAD. Stress falls on PREview, FAILS, WRITE, DOWN, JOB, NUMber, TRY, UPload and aGAIN.
Audio in these samples uses AI-generated voices.
Our free Listen and Repeat practice page has all 6 sets (42 questions) from the same bank as the examples above, with answer checking and an explanation for every option. For a short mixed session each day, try the Speaking questions of the day.
The strategies on this page come from the Listen and Repeat module of the Test Ninjas TOEFL course. Each lesson teaches the English behind the task, then drills it:
Ready for the whole course? Every lesson in all twelve task modules, plus full-length practice tests.
Start my free trial →