CELPIP-General is a single computer-based sitting of about three hours, always in the same order: Listening, then Reading, then Writing, then Speaking. Everything happens at one workstation - audio through a headset, answers by mouse and keyboard, speaking responses recorded into a microphone. Knowing the mechanics before test day is worth real marks, because several of them (single-play audio, fixed speaking timers) punish people who expected something else.
The test at a glance
| Section | Time | Items | What it is |
|---|---|---|---|
| Listening | 47-55 min | 38 questions | Six audio parts, each played once |
| Reading | ~55 min | 38 questions | Four passage-based parts |
| Writing | 53-60 min | 2 tasks | An email and a survey response, typed |
| Speaking | 15-20 min | 8 tasks | Recorded responses to on-screen prompts |
Listening: one play, no going back
Each recording plays exactly once, and once you answer, the test moves forward - there is no replay and no returning to a previous question. The parts cover problem solving, daily-life conversations, information, news items, discussions and viewpoints, progressing from short exchanges to longer, denser audio. The skill being tested is as much sustained attention as comprehension: notes are allowed, but new audio keeps playing while you write.
Reading: four parts, reviewable within each
The four reading parts - correspondence, applying a diagram, information and viewpoints - each pair a text with a block of questions. Within a part you can review and change answers freely until its timer runs out, so the working strategy is different from Listening: skim, answer what is certain, and circle back. The passages are everyday Canadian material - emails, notices, articles - not academic texts.
Writing: two typed tasks
Task 1 is an email (about 27 minutes); Task 2 is a response to a survey question (about 26 minutes). Both ask for 150-200 words, typed, with a live word counter and basic spell-check. Register matters: the prompt tells you who you are writing to, and matching that tone is part of the score. Full annotated examples of both tasks are in our writing samples, and the reusable structures in the free answer templates.
Speaking: eight tasks against fixed timers
Each of the eight speaking tasks - giving advice, describing a scene, making predictions, persuading, and so on - gives fixed preparation time (30-60 seconds) and a fixed response window (usually 60-90 seconds), then records in one take. There is no examiner and no re-recording; the clock drives everything. This is the section where practising under real timers changes scores the most, because pacing a 90-second answer is a learned physical skill, not a language one.
Mechanics worth knowing in advance
- No scheduled breaks - the ~3 hours run continuously. Eat beforehand.
- Unscored experimental items are mixed in and indistinguishable; treat every question as real.
- Scores are 1-12 per skill, aligned one-to-one with CLB levels - the full conversion, including IELTS equivalents, is in the score chart.
- On-screen timers show remaining time per section or task; the test advances itself when they expire.
- Test day logistics - arrival cutoffs, ID rules, what you can bring - are in the booking guide.
The format stops feeling like an opponent after one full rehearsal. A complete mock test here runs the real section order, task types and timers end to end, and scores each skill on the CLB-aligned scale - the closest thing to test day without paying for it.