Waiting a long time for something
A wait you really cared about, and why it mattered. The model lands on an anticlimax and a physical detail rather than a triumph, which is why it works.
Practise this questionFive untimed CELPIP Speaking Task 2 questions, each with a spoken model answer. One story, sixty seconds, and a reflection at the end that most candidates never reach. Record yourself, reveal the model, then read exactly how the story is built.
Task 2 asks for one event out of your own life. A time you waited a long time for something. A time you made a mistake and had to put it right. A time you spoke in front of a group. A time you went back somewhere after years away. A time somebody else made you proud. The prompt gives you a category, and everything inside it, the people, the place, the year, the ending, is yours to supply.
Thirty seconds to prepare and 60 seconds to speak. That halving of the recording time is the whole difference between this task and the one before it, and it changes the answer completely. Sixty seconds is about 130 to 150 words. That is room for one story told properly. It is not room for two stories, and it is certainly not room for the three that nervous candidates reach for when they cannot decide which one is best.
Notice that every Task 2 prompt is actually two questions wearing one coat. "Tell your listener what happened and why it mattered to you." "Describe what happened and how you felt about it." "Tell your listener what happened and why you still remember it." The second clause is not a polite closing formula. It is a scored half of the task, and it is the half most answers never reach, because narrating the events uses up all 60 seconds when nobody has planned the ending.
One liberating fact: none of it has to be true. Nobody can verify your story, nobody tries, and no rater has ever docked a mark for fiction. What they are listening to is whether you can sustain past narration accurately, whether you can select one telling detail instead of piling on adjectives, and whether you can climb from the events to something abstract at the end. If the true version of your story has no turn in it, invent one that does. Choose the version with the most to say, not the most faithful.
Every model answer on this page follows this arc. It is not a formula to recite, it is where the time actually goes when a story works.
Guard those last 15 seconds the way you would guard money. The single most reliable improvement available on Task 2 is deciding your final two sentences during the preparation, before you know how the middle will go. Candidates who plan forwards run out of clock and end on "so, yeah, that was it". Candidates who plan the ending first arrive at it.
The 2026 CELPIP speaking format for the personal experience task specifically.
The story is a delivery mechanism. What arrives inside it is what gets marked, and these are the five things worth carrying.
There is one comparison worth sitting with, because it reframes this whole task. In the published performance standards, the CLB 5 candidate and the CLB 12 candidate are responding to the same material. The gap between them is not what they noticed or how interesting their lives are. It is the descriptive machinery they had available to render it. That is genuinely good news: you cannot make your past more eventful before test day, but you can absolutely arrive with better machinery. Precision. Hedging. A complex clause you can build without thinking. An ending you planned.
Which is also why the "my life is boring" worry is misplaced. A 60 second story about a permit application scores higher than a badly told story about surviving an avalanche. The events are not the material. The telling is.
Five of the commonest experience categories on the test. Pick one, tell the story out loud with nothing counting down, then play yourself back before you unlock the spoken model and the notes on why it lands. The recording never leaves your browser, and nothing here is marked.
A wait you really cared about, and why it mattered. The model lands on an anticlimax and a physical detail rather than a triumph, which is why it works.
Practise this questionThe prompt asks about the fix, not the error, and most answers get that backwards. The model admits the ugly first impulse before saying what it actually did.
Practise this questionThe prompt asks how you felt, and the trap is answering with feeling words. The model shows the nerves through shaking hands and never once says nervous.
Practise this questionMemory against reality, and the whole story exists to earn one abstract line at the end. The easiest of the five to describe like a travel brochure and lose.
Practise this questionThe one story here that is not yours, so the reflection has to come back to you or half the prompt goes unanswered. The model refuses a tidy ending, and the pride outlives the result.
Practise this questionA mistake, a wait, a first time, a return, a moment of nerves. Those five bend to cover almost any Task 2 prompt with a sentence of adaptation. Fifty half-remembered stories give you a decision to make in 30 seconds, which is the one thing you cannot afford.
Banking stories only works if you bend them to the prompt. A memorised answer forced onto the wrong question is audible, and relevance sits at the top of the Task Fulfillment list. Keep the events, change the reflection to answer what was actually asked.
Decide your closing reflection in the prep, before the middle exists. It is the only part of the answer that cannot be improvised well, and it is the part the clock eats. Everything else you can find as you go.
"The one that comes to mind is ..." and you are in. Fifteen seconds of "let me think, there are so many, maybe I could talk about ..." is a quarter of your answer spent proving you did not use the preparation.
"And it went better than the version I had rehearsed." "The hard part was not the waiting." A story a listener enjoys has one moment where the expected thing does not happen. Look for it during prep; if it is not there, choose a different story.
Record one story and check where 60 seconds actually lands in your speech. Most people are shocked by how much fits. After that, practise the shape rather than the clock, which is what this untimed course is for.
CELPIP Speaking Task 2 is Talking about a Personal Experience. A short prompt asks you to describe one event from your own life, such as a time you waited a long time for something or a time you made a mistake. You get 30 seconds to prepare and 60 seconds to speak. It is the second of the eight speaking tasks in the 2026 CELPIP format.
Thirty seconds of preparation and 60 seconds of recording. Sixty seconds is about 130 to 150 words at a natural pace, which is room for exactly one story told properly. It is not room for three stories summarized.
No. Nobody can verify it and nobody tries. Raters score the language, not your biography. Choose the version of events that gives you the most to say, with a clear turn in the middle and something worth reflecting on at the end. A true story with no shape scores lower than an invented one with a good one.
Name the event in the first sentence, anchor it with a date or a duration, tell what happened with one concrete detail rather than a list of adjectives, land the resolution, then spend the last 10 to 15 seconds on why it mattered to you. That final reflection is a scored half of the task, not a decoration.
Most often by finishing the events and stopping. Every Task 2 prompt has two halves: what happened, and why it mattered or how you felt. An answer that narrates the events perfectly and never reflects has answered half the question, and Task Fulfillment scores completeness.
Past tense for the events, present for the reflection at the end. A past experience narrated entirely in the present is a control problem under Listenability, which scores grammar and sentence structure. Sustained, accurate past narration with one shift into the present at the close is exactly what the top bands sound like.
No. Task 2 is text only. A short written prompt appears on the screen during your 30 seconds of preparation and stays there while you speak. It gives you a category of experience rather than a scene, and the story is entirely yours to supply.
Yes, and it is the best use of your preparation time overall. Four or five flexible stories, a mistake, a wait, a first time, a return, a moment of nerves, will cover most prompts with light adaptation. Just make sure you adapt them: a memorised story bent to fit the wrong prompt is obvious, and relevance is scored.
No. There is no timer and no evaluation anywhere in the practice course. You record yourself so that you can hear yourself, the recording stays in your browser and is never uploaded, and then you reveal a spoken model answer with its structure, key phrases and common mistakes.
Yes. Everything here follows the 2026 CELPIP speaking format: eight tasks, a timer that runs per task and resets on every advance, and Task 2 running 30 seconds of preparation followed by 60 seconds of recorded speech.
Task 3 takes your own life out of it entirely and hands you a scene to describe. Or see every model answer on one page.
Describe a picture to somebody who cannot see it. Actions, not objects, and a route through the room.
Learn Task 3 → 1Ninety seconds on somebody else's problem, where the tone is chosen for you by the relationship.
Learn Task 1 →Every task, every question, every model answer, on one page with the topics listed in full.
See the samples →All 8 tasks back to back with the real prep and response clocks running.
See the speaking section →Five experience questions, no clock, no score. Record your story, listen to where it sags, then compare it with a model that spends its last fifteen seconds where the marks are.