CELPIP Speaking Task 4: Making Predictions (Sample Answers + Future Forms)

1What CELPIP Speaking Task 4 is

Task 4 of the CELPIP Speaking test is called Making Predictions. You are shown the exact same picture you just described in Task 3, and your job changes completely: instead of saying what is happening in the scene, you predict what is going to happen next. You imagine the next few moments — the minutes that follow the frozen image — and you say them out loud.

The catch is that your predictions cannot be random guesses. A strong Task 4 answer is anchored to the picture: every prediction grows out of a visual clue you can actually point to. If a child is reaching for a balloon, you predict the balloon will slip away. If the sky is bright and clear, you predict the market will get busier as the morning goes on. You are reading the scene like a story and saying what comes on the next page.

So Task 4 is really two skills in one. First, the language of the future — will, going to, might, is likely to — used accurately and with variety. Second, the reasoning that connects what you see to what you predict, so the rater hears a logical mind, not a lucky guesser.

Task 4 is not describing the picture again. It is predicting the next 60 seconds of the scene and justifying every prediction from a clue you can see in the image.

2Timing & format

Task 4 follows immediately after Task 3 and uses the same picture, so there is no new image to study from scratch — you already know the scene. The timing is short and fixed, and the screen tells you exactly where you are in the clock.

Preparation time: 30 seconds. The prompt and the picture appear, and you plan silently. No notes are scored, but you can think and organise your predictions.

Speaking time: 60 seconds. The microphone opens automatically and records until time runs out. There is no early stop and no second take.

Same image as Task 3: you predict what happens next in the picture you have just spent Task 3 describing — the scene continues, it does not change.

Prompt wording: the instruction always asks you to predict, for example 'Predict what is going to happen next' — that verb tells you to look forward, not to describe.

Thirty seconds is barely enough to write words down, so spend the prep choosing two or three things you will predict and the clue behind each one. Sixty seconds of speaking is room for an opening, three predictions with reasons, and a short closing line — no more. Plan for that shape and you will never run out of things to say or trail off into silence.

Don't waste prep time writing full sentences. Jot three quick triggers — 'kids → balloon go', 'clouds → none → busier', 'vendor → sell out' — and build the sentences live. That is faster and sounds more natural than reading.

3How Task 4 is scored

Like every CELPIP Speaking task, Task 4 is rated on the same four dimensions. Knowing how each one shows up in a prediction answer tells you exactly what the rater is listening for.

Content & Coherence: Are your predictions clear, logical, and connected to the picture? A high score needs predictions that follow from visible clues and flow in a sensible order, not a scattered list.

Vocabulary: Do you use precise, varied words to talk about the scene and the future? Repeating 'happen' and 'thing' caps your band; specific nouns and verbs lift it.

Listenability: Can the rater follow you easily — steady pace, clear pronunciation, natural rhythm, no long pauses while you hunt for 'will'?

Task Fulfilment: Did you actually make predictions about what comes next, with justification — not slip back into describing the picture?

Two things matter most for the band on this particular task. The first is varied future forms: an answer built only on 'will' sounds flat and limited, while a mix of going to, might, could, and is likely to shows control of the grammar of the future. The second is logical prediction from clues: the rater wants to hear the bridge between what you see and what you predict — 'Since the sky is clear, the market will probably get busier.'

You don't have to be right. There is no correct future. You are scored on how well you predict and justify in English — a wild but well-reasoned prediction beats a safe one with no reason behind it.

4The winning structure for Task 4

A reliable five-move shape fits perfectly into 60 seconds and hits every scoring dimension. Learn it once and reuse it on any picture, so on test day you spend your energy on ideas, not on wondering what to say next.

Open with the most obvious next event. Name the single thing that will clearly happen next so you start fast and confident: 'In the next few minutes, the market is going to get even busier.'

Predict two or three specific things. Move from the obvious opening to concrete, named predictions about particular people or objects in the scene — not vague generalities.

Justify each one from a visual clue. After every prediction, give the reason you can see: 'because the stalls are already crowded' or 'since several people are carrying empty bags.'

Vary your certainty. Don't make every prediction equally sure. Mix strong forms (is going to, will definitely) with tentative ones (might, could, is likely to) so your grammar and your judgement both show range.

Close with an overall prediction. End with one summarising line that ties the scene together: 'Overall, it looks like it's going to be a busy, successful morning for the market.'

Memorise the five moves, not a script. The structure stays the same on every picture; only the clues and predictions change. That is what keeps you fluent under a 30-second clock.

5Future forms & prediction language

Variety in how you talk about the future is one of the fastest ways to lift your Vocabulary and Content scores on Task 4. Pull from each of the four groups below so no two predictions sound the same.

Certainty — high confidence

Use these for the obvious, near-certain next events: 'is going to', 'will definitely', 'is bound to', 'is about to'. Example: 'The vendor is about to hand over the change.'

Possibility — lower confidence

Use these for predictions you are less sure about, so your judgement sounds nuanced: 'might', 'could', 'may', 'is likely to', 'is probably going to'. Example: 'The child might drop the apple.'

Reasoning — connect clue to prediction

These phrases build the bridge the rater wants to hear: 'Judging by…', 'Since the sky is clear, …', 'Given how crowded it is, …', 'Because of…', 'That suggests…'. They turn a guess into a logical prediction.

Sequencing — order the future

Words that place predictions in time make your answer flow: 'In a few minutes…', 'Soon…', 'Later…', 'By the end of the morning…', 'After that…'. They stop your predictions sounding like a random list.

Aim to use at least three different certainty levels in one answer. Hearing 'is going to', 'might', and 'is likely to' in 60 seconds is exactly the grammatical range that separates CLB 9 from CLB 7.

6Sample answer at CLB 7

Scenario: Look again at the busy farmers' market scene on a sunny morning. Predict what is going to happen next. This CLB 7 answer is clear and on-task, leans mostly on 'will' and 'going to', and gives simple reasons. It does the job, but the future forms and reasoning are basic.

In the next few minutes, the market is going to get more busy. I think more people will come because it is a sunny morning and many people like shopping then. The woman near the fruit stand will probably buy some apples, because she is looking at them and holding a bag. The vendor will sell a lot of vegetables today. The two children will run around the market, because children are always excited in busy places. The man with the dog will keep walking and maybe he will stop to buy something. I think the stalls will sell many things and they will be happy. So I think it is going to be a good and busy morning for everyone at the market.

Why CLB 7: The answer is clear, on-task, and makes several predictions with simple reasons. But almost every prediction uses 'will' or 'going to', the vocabulary repeats ('busy', 'buy', 'many'), and the justifications are thin. It is fully understandable — that is solid CLB 7 — but it lacks the range of future forms and the precise reasoning that push into CLB 9.

7Sample answer at CLB 9

Scenario: Look again at the busy farmers' market scene on a sunny morning. Predict what is going to happen next. Same picture, clearly stronger answer. Notice the mix of future forms, the clearer link from clue to prediction, and the smoother flow.

Judging by how crowded the stalls already are, the market is going to get even busier over the next hour. Since the sky is completely clear, more shoppers are likely to arrive, so the popular stalls will probably start selling out of their best produce. The woman examining the apples is about to fill her bag and pay, because she's already holding one up to check it. The vendor might run low on tomatoes soon, given how fast that line is moving. The two children could knock something over if they keep running between the tables — they look a little too excited. Later in the morning, some vendors will likely restock or drop their prices to clear what's left. Overall, judging by the weather and the crowd, it's going to be a successful, profitable morning for almost everyone here.

Why CLB 9: The future forms vary — 'is going to', 'are likely to', 'is about to', 'might', 'could', 'will likely'. Every prediction is tied to a visible clue with a reasoning phrase ('Judging by…', 'Since…', 'given how…'). The vocabulary is more precise ('produce', 'restock', 'profitable') and the sequencing words give it flow. That range and control is the CLB 9 zone.

8Sample answer at CLB 11

Scenario: Look again at the busy farmers' market scene on a sunny morning. Predict what is going to happen next. This top-band answer adds nuanced certainty levels, precise reasoning, and the easy, natural flow of a confident speaker.

Given how packed the stalls already are this early, the market is almost certainly going to hit its peak within the hour. With the sky this clear, a steady stream of shoppers is likely to keep arriving, which means the busiest vendors will probably sell out of their freshest produce well before noon. The woman inspecting the apples is on the verge of buying — she's weighing one in her hand, so a purchase looks imminent. The tomato seller, judging by that fast-moving queue, may well need to restock or risk turning customers away. I'd be a little wary about those two children darting between the tables; if no one slows them down, they could easily send a crate tumbling. As the morning wears on, I'd expect the later stallholders to start discounting whatever hasn't sold, just to clear it. All things considered, this looks set to be a brisk, profitable morning, with the weather doing most of the work.

Why CLB 11: The certainty is finely graded — 'almost certainly', 'is likely to', 'may well', 'could easily', 'I'd expect' — so the speaker sounds genuinely thoughtful, not formulaic. The reasoning is precise and embedded naturally, the idiom is native-like ('on the verge of', 'looks imminent', 'looks set to be'), and the whole answer flows as one connected argument. That nuance and ease is the CLB 11 ceiling.

9Common mistakes on Task 4

Most lost marks on Task 4 come from a handful of repeat errors. Catch yourself on these and your band rises without learning a single new word.

Describing the picture again instead of predicting. The biggest trap: Task 4 reuses the Task 3 image, so it's easy to slip back into 'There is a woman buying apples.' That is description, not prediction — say what will happen next, not what is happening now.

Using only 'will'. An answer built entirely on 'will' sounds flat and caps your grammar score. Mix in 'going to', 'might', 'could', and 'is likely to' to show real control of the future.

Predictions with no justification. 'The market will get busier' is half an answer. Without 'because the stalls are already crowded', the rater can't hear your reasoning, and reasoning is exactly what this task tests.

Unrealistic predictions. Saying a spaceship will land or the market will catch fire breaks the logic the picture supports. Keep predictions plausible and grounded in what the scene actually shows.

Forgetting it follows Task 3. Treating Task 4 as a fresh prompt wastes time re-orienting. Remember it continues the same scene — your Task 3 observations are the raw material for your Task 4 predictions.

10How to practise Task 4

Task 4 is a trainable skill, not a talent. A short daily routine with any photo builds the reflex of predicting and justifying under time pressure.

Take any photo and predict the next 60 seconds. Use a magazine image, a phone snapshot, or a stock photo. Look at it and ask: what happens in the next minute? Say it out loud.

Vary your future forms deliberately. Force yourself to use at least three different ones in each practice answer — 'is going to', 'might', 'is likely to' — so variety becomes automatic, not something you reach for.

Justify every prediction from a clue. Train the bridge: after each prediction, add 'because…' or 'since…' and name what in the image makes you think so. No prediction without a reason.

Time it at 30 seconds prep and 60 seconds speaking. Use a real clock so the pressure matches the test. Plan in 30, speak for the full 60 — never stop early.

Record yourself. Use your phone so you can hear what the rater hears — your pace, your pauses, your filler words. You can't fix what you can't hear.

Review and redo. Listen back, note one weakness (too much 'will', a missing reason, a long pause), then redo the same picture once with that one fix. Small targeted reps beat endless new prompts.

Practising the same photo twice — once raw, once after a review note — improves you faster than chasing a new image every time. The second attempt is where the learning sticks.

11How FlexiLingo helps you master CELPIP Speaking

FlexiLingo turns this practice routine into a guided loop, so you get the reps, the feedback, and the model language without needing a tutor on call.

AI speaking practice on CELPIP-style prompts

Practise Making Predictions on picture prompts built to mirror the real Task 4, with the 30-second prep and 60-second speaking clock so the pressure matches test day.

Instant feedback

Get immediate notes on your future forms, your reasoning, your pace, and your filler words right after you speak — the same dimensions a CELPIP rater listens for.

Model answers by band

See sample responses at CLB 7, 9, and 11 for the same prompt, so you can hear exactly what a higher band sounds like and copy the moves that lift your score.

Vocabulary in context

Save the prediction language and precise nouns you meet — 'is likely to', 'on the verge of', 'restock' — with their full sentence as context, straight into your review deck.

Spaced-repetition review

The future forms and vocabulary you collect come back at the optimal moment, so the language you need on test day is automatic, not something you have to search for.

Frequently Asked Questions

How is Task 4 different from Task 3?

They use the same picture, but the job is opposite. Task 3 (Describing a Scene) asks you to describe what is happening right now in the image. Task 4 (Making Predictions) asks you to predict what will happen next — the moments after the frozen scene. Task 3 lives in the present; Task 4 lives in the future. The single most common Task 4 mistake is slipping back into description, so the moment you start, switch your verbs to the future and keep them there.

Which future forms should I use?

Use a mix, not just 'will'. Combine certainty forms ('is going to', 'will definitely', 'is about to') with possibility forms ('might', 'could', 'is likely to', 'may'). Add reasoning phrases ('Judging by…', 'Since the sky is clear, …') and sequencing words ('soon', 'later', 'by the end of the morning'). Hearing three or four different future forms in one 60-second answer is exactly the range that pushes you from CLB 7 toward CLB 9 and above.

Do my predictions need to be realistic?

Yes, plausible — but there is no single correct answer. You are not scored on whether your prediction comes true; you are scored on your English and your reasoning. Predictions should be grounded in what the picture actually shows: if a child is reaching for a balloon, predicting the balloon slips away is logical. Predicting a spaceship lands breaks the logic the image supports and works against your Content score. Stay realistic and tie every prediction to a visible clue.

Sixty seconds — how many predictions should I make?

Two or three solid predictions, each with a justification, fits 60 seconds perfectly. That leaves room for a quick opening line and a short closing summary. More than three and you'll rush, drop your reasons, or run out of time mid-sentence. Quality beats quantity here: three predictions, each tied to a clue and phrased with a varied future form, scores far higher than five rushed guesses with no reasoning.

How do I reach CLB 9 on Task 4?

Three things lift you to CLB 9. First, vary your future forms — get at least three different ones into the answer instead of leaning on 'will'. Second, justify every prediction from a visible clue using a reasoning phrase, so the rater hears the logic, not just the guess. Third, keep it flowing — use sequencing words and a clear opening and closing so the answer sounds like one connected idea, not a scattered list. Drill those three with the 30/60 clock until they're automatic.