Items
About 7 sentences
Response time
About 8 to 12 seconds each
Replays
None
Setting
Often a guided tour or a campus scene

What the task looks like

A short scene sets the context, such as a guide showing new students around a library. You then hear one sentence at a time. After each one the recording starts and you repeat it. Early sentences are around six words; later ones can pass fifteen.

How it is scored

Each answer is judged on how closely it matches the original and how clearly you say it. Missing or changing words lowers the score, and so does unclear pronunciation. A complete sentence with one small slip scores far better than a fast, mumbled one.

A four-step method

  1. 01Listen for meaning, not for words. If you understand the idea, your brain can rebuild the grammar.
  2. 02Split the sentence into meaning groups in your head: who, did what, where, when.
  3. 03Start speaking as soon as the recording begins. Waiting lets the sentence fade.
  4. 04Copy the speaker’s rhythm and stress. Rhythm carries the words back into memory.
Splitting a long sentence

The study rooms on the second floor | can be reserved online | for up to three hours a day.

Answer

Three groups instead of eighteen separate words.

Each group is a unit of meaning. Remembering three ideas is much easier than remembering a long string of words.

Common mistakes

  • Dropping small words: "Library opens at nine"
    The library opens at nine.
  • Changing the tense: "The tour starts at ten"
    The tour started at ten.
  • Swallowing endings: "She work in the lab"
    She works in the lab.
  • Stopping to restart halfway through
    Keep going and fix nothing.

Daily drills

  • Shadowing: play a short clip and speak along with it, one second behind.
  • Sentence ladders: repeat a 6-word sentence, then 9, then 12, then 15.
  • Record and compare: write down what you said and check it word by word.

Try Listen and Repeat with a real timer in a free full test.

Start free

Your band estimate is one free test away.

One full timed test, an estimated band and AI feedback. No card needed.