Do I have to describe everything in the picture?
No, and trying to will cost you points. Choose four or five clear elements, give each one a location, an appearance detail, and an action, and let the rest of the picture go.
Back to questions
Speaking Task 3Demo
Describing a Scene
The exact exam interface with a sample question. Nothing is recorded or saved.
CELPIP Speaking (Demo) - Task 3 1/1
Preparation time: 30 seconds
Time remaining: 1 minute
Describe some things that are happening in the picture below as well as you can. The person with whom you are speaking cannot see the picture.
Preparation Time
30
Step 1 of 3: preparation time. Scan the picture for the parts worth describing, the people, actions, and places, and plan a route through them. In practice you can skip this wait. In the real exam you cannot, and it ends with a loud "Start speaking now".
CELPIP Speaking Task 3 Guide:Describing a Scene Tips, Template and CLB 9 Sample
In Task 3 the screen shows a detailed picture of an everyday scene, and your job is to describe it to someone who cannot see it. You get 30 seconds to prepare and 60 seconds to speak. Raters are not testing your imagination, they are testing whether your words alone let a listener rebuild the scene. The demo above runs on the exact interface and timing of the real test, so everything you practice here transfers directly.
The format at a glance
Preparation
30 seconds to scan the scene and pick four or five things worth describing
Response
60 seconds of recorded speech, the timer submits automatically
Listener
Someone who cannot see the picture, so every detail needs a spoken location
How Task 3 is scored
Trained raters score every CELPIP speaking task on the same four equally weighted dimensions. In Task 3 each one rewards a specific describing skill, which makes the points unusually easy to aim at.
Content and Coherence
Whether your details add up to one scene the listener can follow. A single clear route through the picture scores better than a pile of unconnected observations.
Vocabulary
Range and precision. Spatial phrases like in the foreground plus exact verbs like reaching and chatting show more range than any rare idiom would.
Listenability
Rhythm, pronunciation, intonation, and grammar working together. In Task 3 that mostly means keeping present continuous steady, a man is climbing, not a man climb.
Task Fulfillment
Did you actually describe this picture, in enough spoken detail that a listener could sketch it? Drifting into opinions about the topic loses more points here than any grammar slip.
A 60 second description template that works
Preparation30 seconds, before the answer
Scan the picture and choose a route: the overall setting first, then four or five people or groups to cover along the way. When the 30 seconds run out you hear a loud "Start speaking now" and recording begins immediately, there is no button to press.
Open with the big pictureabout 10 seconds
Name the kind of place and the overall feel in one or two sentences, like a bright, busy public library. That frame lets the listener place every detail that follows.
Walk one route through the scene35 to 40 seconds
Move in one direction, foreground to background or left to right, and give each person a location, an appearance detail, and an action in present continuous. A fresh spatial phrase signals every move.
Close on a detail or the mood10 to 15 seconds
End with one small specific detail or a sentence about the atmosphere so the answer sounds finished on purpose. Getting cut off mid-sentence by the timer reads as running out of language.
Location phrases that guide your listener
Each phrase tells your listener where to look before you say what is happening there. Open every new part of the description with one and the scene builds itself in order.
On the left side of the picture...In the foreground, I can see...Right behind them...In the background, there is...
Mistakes that cost points
Jumping around the pictureContent and Coherence
Describing whatever you notice next forces the listener to rebuild the scene after every sentence. Pick one route through the picture and follow it.
Slipping out of present continuousListenability
A man climb, or a sudden jump into past tense, is the easiest error for raters to hear. The actions are happening now, so keep them in is doing form.
Vague words like thing and stuffVocabulary
Some people are doing something gives the rater nothing to score. Name the clothing, the object, and the exact action instead.
Trying to describe everythingContent and Coherence
Racing through a dozen items leaves no room for detail and your delivery falls apart. Four or five well-described elements beat a complete inventory.
Pointing at what the listener cannot seeTask Fulfillment
This person or over here means nothing to someone without the picture. Anchor every detail with a spoken location like on the left or near the back.
Task 3 questions people ask
What if I don't know the word for something?
Describe around it and keep talking. A tool for reaching the top shelves works fine when ladder will not come to mind, and raters reward a smooth recovery far more than they punish a missing word.
Is the Task 4 picture the same as this one?
Yes, Task 4 shows the same picture and asks what will happen next. While you describe the scene here, quietly note one action in progress, like someone reaching for a book, and your Task 4 prediction is already made.
