PracticeCLB
CELPIP Listening Course
Lesson 18 min

Understanding the CELPIP Listening Format

Six parts, one session

CELPIP Listening has 6 parts, each with multiple questions. You hear each audio clip once — there is no replay.

PartNameFormat
1Listening to Problem Solving5 questions on a conversation about solving a problem
2Listening to a Daily Life Conversation5 questions on an everyday conversation
3Listening for Information5 questions on an informational dialogue
4Listening to a News Item5 questions on a radio or podcast-style report
5Listening to a Discussion8 questions on a multi-speaker discussion
6Listening to Viewpoints8 questions on a debate or opinion exchange

Total: 36 questions across 47–55 minutes.


The single most important Listening skill: pre-reading questions

Unlike most other standardised tests, CELPIP Listening gives you time to read each question before the audio plays. Most candidates ignore this and just wait for the audio to start.

This is a mistake. Pre-reading the question tells you:

  1. What to listen for — you're not listening to everything equally. You're hunting for a specific answer.
  2. The type of information the question targets — a "why" question tells you to listen for reasons. A "what" question tells you to listen for a noun/event. A "how" question tells you to listen for a process.

Pre-reading technique:

  • Scan all 5 questions before the audio starts
  • Identify the key word in each question (the word that tells you what to listen for)
  • Ignore the answer options until you've heard the relevant part of the audio

Why note-taking matters for Parts 5 and 6

Parts 5 and 6 are significantly longer than Parts 1–4. The discussion format means information is distributed across multiple speakers, and questions often require you to identify who said what or synthesise across the conversation.

Minimal note format that works:

  • One column per speaker (or label S1, S2)
  • Bullet the key point from each speaker turn
  • Circle anything that sounds like a direct answer to a question you pre-read

You don't need to write full sentences. You need a map of who said what so you can answer accurately without trusting memory alone.