All lessons

Module 5: The eight speaking tasks · Lesson 5.5 · about 18 min

Speaking Task 4: Making Predictions

CELPIP Speaking Task 4: say what will happen next in the Task 3 picture in 60 seconds. Clue-based predictions, future forms, two models and 28 drills.

Key points, meanings and formulas on one page: choose "Save as PDF" in the print box.

Module 5 · about 18 minutes. By the end you will be able to turn clues in a picture into believable predictions, use the future forms that make them sound natural, compare two model answers and practise with three scenes. You will also answer 28 practice questions.

Task 4 is detective work: every prediction starts with a clue you can see. Point to the clue, say what will probably happen, and give the reason, then chain one event to the next.

1. What Task 4 looks like

Task 4, Making Predictions, shows you the same picture as Task 3. This time you say what will probably happen next, usually in the next few minutes.

  • You usually get about 30 seconds to prepare and 60 seconds to speak. The test screen shows the exact time.
  • Because you have just described the picture, you already know it well. The 30 seconds are for choosing your predictions, not for looking at the picture again from scratch.
  • The future is open, so there are no wrong predictions, only weak ones (no clue, no reason, or something unbelievable).

Everyday analogy. A weather forecaster does not guess at random. She sees clouds moving in, wind picking up and pressure falling, and says “rain is likely this afternoon”. You do the same with the picture: clouds, signs, someone not looking, a tired worker. Each clue becomes a forecast.

2. What it really tests

  • Content/Coherence: 3 or 4 predictions, each tied to a clue and a reason, in a clear order, ideally linked (if this happens, then…).
  • Vocabulary: words for consequences and reactions (trip, spill, slide, scramble, take shelter, hand out a ticket).
  • Listenability: future forms used correctly and with variety (will, be going to, might, is likely to, if… will), plus linking words for cause and effect.
  • Task Fulfillment: talk about the future, not a second description; cover several people or areas; use most of the 60 seconds.

3. The method: Clue, Prediction, Reason

Prep (30 s): circle 3 or 4 clues. From your Task 3 description, pick the details that suggest change: weather, signs, people not paying attention, things about to fall, lines, clocks. Note them with arrows: clouds → rain, cover tables / girl phone + step → trips, eggs / boy → ice cream / rain → guitar case.

Opening (about 8 s): the big forecast. One sentence on the overall future: “I think the next few minutes at this market are going to get a lot more hectic.”

Predictions (about 40 s): Clue, Prediction, Reason, for each. “The girl in the middle is staring at her phone (clue), so she will probably trip on the step (prediction), and a few of those eggs are likely to end up on the ground (result).”

Chain events with if. One prediction can cause the next: “If it starts to rain, the musician will have to close his guitar case quickly.” Chains fill time naturally and show complex grammar.

Close (about 8 s): sum up. “All in all, I think the morning will end with a rush to get out of the rain, but most people will still go home with their shopping.”

4. Worked example: the farmers’ market, next

The same picture as the lesson 5.4 example: a sunny farmers’ market at 10:15, dark clouds behind the clock tower; a woman weighing apples for an older man; a boy with a balloon pulling his father toward an ice cream stand with a line; a guitarist with an open case and a sleeping dog; a teenage girl carrying a tall stack of egg cartons, looking at her phone, walking toward a step.

Thinking in prep: clues for change are (1) the clouds, (2) the girl, the phone and the step, (3) the boy and the ice cream line. I can chain rain → sellers cover tables → musician packs up. Order: girl first (the most immediate), then the boy, then the rain chain, then a close.

Model answer, CLB 7 level (about 100 words)

I think some things will happen soon in this picture. First, the girl with the eggs is looking at her phone. She will probably fall on the step and drop the eggs. Then she will be sad. Second, the little boy wants ice cream. I think his father will buy it for him, but they will wait in the line. Also, there are dark clouds, so it will rain soon. The sellers will cover their tables. The man with the guitar will close his case. In the end, many people will go home because of the rain.

Why it scores around this level. Content/Coherence: clear predictions, each with a clue, and simple order words (first, second, also). Vocabulary: clear but basic (sad, fall, go home). Listenability: “will” in almost every sentence; correct but not varied. Task Fulfillment: stays in the future, covers several areas; slightly short.

Model answer, CLB 10 level (about 145 words)

Judging by a few things in this picture, I think the next few minutes are going to get a lot more hectic.

The most immediate thing is the girl in the middle. She is staring at her phone instead of watching where she is going, so she will almost certainly catch her foot on that step, and at least a couple of those egg cartons are likely to hit the ground.

Meanwhile, the little boy will probably win the argument with his dad, although they will have to queue for a few minutes before he gets his ice cream.

The bigger change, though, is the weather. Those clouds behind the clock tower are moving in, so I expect a shower fairly soon. If it starts to rain, the stallholders will scramble to cover their produce, and the musician will have to shut his case before his tips get soaked.

All in all, a sunny morning that ends in a rush for shelter.

Why it scores higher. Content/Coherence: every prediction names its clue (staring at her phone, clouds moving in), and the answer moves from the most immediate event to the biggest change, then sums up. Vocabulary: precise and natural (hectic, catch her foot, queue, scramble, produce, soaked, rush for shelter). Listenability: varied future forms (are going to, will almost certainly, are likely to, will probably, I expect, if… will have to) and linking (meanwhile, although, though). Task Fulfillment: entirely about the future, several areas covered, about 60 seconds.

5. Traps in Task 4

Trap Example Fix
Describing again “A woman is weighing apples…” Use the present only as a clue: “Because…, she will…”
No reason “The man will leave.” Add the clue: “…because it is starting to rain.”
Too dramatic “A tornado will destroy the market.” Keep it likely and small: a shower, a trip, a ticket
“Will” every time “He will… she will… they will…” Mix: is going to, might, is likely to, I expect, if… will
One person only 60 seconds on the girl 3 or 4 people or areas
Too far in the future “In ten years, she will be a doctor.” The next few minutes or hours

Self-check after every Task 4 practice:

  • I opened with an overall forecast for the scene.
  • Every prediction was linked to a clue I could see, with because or so.
  • I talked about the future, not about what is happening now.
  • I used at least three different future forms (will, be going to, might, is likely to, if… will).
  • I linked at least two predictions with if or as a result.
  • My predictions were believable, about the next few minutes or hours.
  • I covered 3 or 4 people or areas and closed with a summary.

“The more surprising my prediction, the higher my score.” A surprising prediction with no clue sounds random, and it makes your answer harder to follow. Raters reward predictions that are clear, supported and well expressed. “Because she is not watching where she is going, she will probably trip” is stronger than “an alien will land in the square”.

  • Opening: Judging by…, I think… ; It looks like things are about to… ; I expect the next few minutes will…
  • Sure: will almost certainly ; is bound to ; is going to (from what we can see)
  • Likely: will probably ; is likely to ; I expect ; chances are
  • Possible: might ; may ; could ; there is a chance that
  • Cause and effect: because ; so ; as a result ; If…, … will… ; Unless…, … will… ; That means…
  • Closing: All in all… ; In the end, I think… ; By the time…, everyone will…

6. Timed practice: three scenes, what happens next

These are the same scenes as the practice tasks in lesson 5.4. If you have just described one, predict for it now, as in the real test.

Practice scene 1 of 3: moving day. Clues to use: the sign and the officer, the slipping grip, the dripping water.
Practice scene 2 of 3: the skating rink. Clues to use: the snow, the sign, the thin ice, the worker's watch.
Practice scene 3 of 3: the library. Try a chain: the cart, the open door and the stroller.

7. Practice: making predictions

28 questions: the same three scenes as lesson 5.4 (choose the best-supported prediction), future forms, sure and possible language, cause and effect, word order and word stress.
  • Task 4: usually 30 s to prepare, 60 s to speak, about the same picture as Task 3.
  • Every prediction = Clue + Prediction + Reason; start with the most immediate event.
  • Mix future forms: will probably, is going to, might, is likely to, if… will.
  • Chain events with if and as a result to fill time naturally.
  • Stay in the near future, keep predictions believable, and close with a summary.

Check yourself

1What makes a strong Task 4 prediction?
2Which sentence is a well-formed prediction?
3How do Task 3 and Task 4 differ?

Sources

Your ticks stay in this browser. Save them as a file on the course home.

Updated 2026-11-22. fibub is an independent free site, not part of Paragon Testing Enterprises or CELPIP, and every question here is original. Test rules, times and immigration requirements change: check celpip.ca and canada.ca before you book. Practice runs in your browser; nothing you type or record is sent anywhere.