CELPIP Speaking Task 4: Making Predictions
Say what might happen next in a scene and connect each prediction to something visible.
At a glance
- Preparation
- 30 seconds
- Speaking
- 60 seconds
- Response
- 1 recorded task
Preparation and speaking have separate timers. Read the role and task carefully, then finish your response within the recording time.
Speaking lasts about 15 minutes overall. There is no prescribed word count; organize enough relevant detail to use the speaking time naturally.
Format reference: official Speaking study pack · current test format. Checked 4 October 2026.
Use your preparation time
Find two or three visible clues that suggest what will happen next. Pair each prediction with the evidence that makes it plausible.
Manage your speaking time
Use the 60 seconds to explain likely next actions and their consequences. Give each prediction a reason and finish the last prediction clearly.
Response plans are MockPip suggestions; follow the official preparation and recording timers.
A technique for this task
Use “is likely to”, “may” or “will probably” to match your certainty. Describe the clue briefly, then explain the next event rather than repeating the scene description.
Try an original practice scenario
Practice focus: Imagine a community-garden picture where someone is carrying a full watering can towards new seedlings.
This example is independently written for practice; it is not a question from an official exam.
Plan a clear response
- Give a likely next action for a person in the scene.
- Support it with a visual clue.
- Add another plausible development without describing the same action twice.
Example of a useful opening
The person carrying the watering can will probably water the new plants next. Since another volunteer is preparing soil nearby, they may plant another row together.
Use this as a model of structure. In your own attempt, respond naturally to the prompt you receive.
A common mistake
Repeating a description of the picture does not answer what is likely to happen next.
Check your recording
Is each prediction linked to a detail you could actually see? Play your response back, choose one thing to improve, and record another attempt. If feedback is available for your attempt, use it alongside what you hear yourself.
Your next timed exercise
Write each prediction beside its visible clue after recording. Replace any unsupported story with a prediction grounded in the image.
Check whether your answer completes the role, uses specific supporting details and ends before the recording timer expires.
MockPip is independent and is not affiliated with or endorsed by CELPIP or its test provider. Practice feedback is not an official test result.