Part 3 is a talk packed with prices, dates and conditions, where one speaker holds the facts the other needs. Play two short clips with no timer, decide your answers, then reveal the transcript and a reason for every option, right down to the figure it borrowed. The distractors are all real numbers, put in the wrong place.
Listening for Information is the part where the exam stops being about people and starts being about facts. Two speakers again, one woman and one man, but now one of them is the person who knows: a garden coordinator, a box office agent, anyone whose job is to explain how a thing works. The other person asks the questions you would ask, and the answers come back thick with numbers, deadlines and conditions.
On the real 2026 test the clip is a little longer than the earlier parts, roughly two to two and a half minutes, and it carries six questions afterwards. What makes it harder is density, not speed. A single sentence from the expert can contain a fee, a date and a catch, and the specific-information questions are built to see whether you kept them apart. The topic is easy to follow; the details are not easy to hold.
The signature trap of Part 3 is that almost every wrong option is a true number said somewhere else in the clip. If the small garden plot costs forty dollars and the big one costs eighty, the question about the small plot will offer you eighty, because you heard it and it feels safe. The test is never asking whether a figure was mentioned. It is asking which fact that figure belongs to.
The two clips below are each an Information exchange in that shape: a person with a need, a person with the answers, and a scatter of prices and dates in between. Learn to file each figure against the right fact here, then take the full six-question version under the clock.
How Listening for Information is built on the timed 2026 test, so you know what these clips are training.
The core move in Part 3 is to write down every number and date you catch, and to tag each one with the fact it describes. A figure with no label is a figure you will hand to the wrong question thirty seconds later.
The tagging habit is what separates Part 3 from a memory game you are bound to lose. Six questions, each anchored to one figure or date, and a clip full of near-identical numbers. Reasoning cannot rescue a figure you filed without a label, so the note-taking is not optional decoration, it is the whole method.
The pairing habit matters because the test writers build their distractors from proximity. Two prices mentioned in the same sentence are a gift to a decoy: whichever one you did not need is dropped into the option list looking perfectly plausible. If your notes already say which price does which job, the pair holds no danger.
Each clip opens in the player. Listen once or as often as you like, answer the questions shown as text, then reveal the transcript and a reason for all four options on each question. Nothing here is timed and nothing is scored.
Warren stops by a community garden in London, Ontario, to ask Adaeze, the garden coordinator, about renting a plot for the summer. Two plot sizes, two fees and a sign-up deadline all arrive close together, and the questions want the right one of each.
Simona asks Neil, the box office agent at the Wortley Playhouse in London, Ontario, about buying a season subscription. A season price sits next to a per-play price, a subscriber perk sits next to a deadline, and each question picks one thread out of the tangle.
On the real 2026 test the Information clip plays once, begins on its own and cannot be paused, and the six questions are then spoken to you one at a time with about thirty seconds each, in order. With this many figures packed into two minutes, that single listen is a genuine test of your note-taking, which is exactly why it is worth rehearsing under those rules.
This course loosens all of that so the facts can teach. Replay the clip until you can hear which number goes where, read the questions as text rather than catching them once, and when you reveal, you see the whole transcript and a written reason for every option, including which figure each wrong one borrowed. Only the clip is audio here; on the real test the Part 3 questions are spoken, so the printed versions are a study aid rather than a copy of the exam.
On the voices: they are North American, because we have no access to a Canadian text-to-speech voice. The content is Canadian, from the London, Ontario, setting to the community garden and the neighbourhood playhouse, so the phrasing you train on is the phrasing you will meet. When you want the timed version, the timed listening tests put all six parts under the real clock.
The specific ways careful listeners still lose marks when the clip is full of facts.
The question asks the small plot's fee and offers the large plot's eighty, because you heard eighty too. A number being real is not the same as it being right. Only the tagged figure answers the question.
Thirty-two per single play and one hundred and fifty for the whole season are different animals. The option that offers the per-play price for a season question is built precisely for the listener who missed the word "season".
Prices feel like the point, so dates slip by. But sign up by April fifteenth and subscriptions ending September thirtieth are questions in their own right, and a listener fixed on money walks straight past them.
Six near-identical numbers in two minutes will not survive in your head. On test day you are handed paper for a reason, and the candidates who do not use it in Part 3 are the ones guessing by the third question.
Only a subscriber can exchange a ticket. Attach that benefit to the wrong person and the option that says single-buyers can do it looks fine. Note which choice earns which advantage.
The correct option paraphrases the clip, so the fee may be described rather than named the same way. A listener holding out for the identical phrase overlooks the answer that reworded it.
Two columns: the thing and its number. Small plot, 40. Large plot, 80. Season, 150. A table keeps figures from bleeding into each other, and when a question names the thing, the number is already sitting beside it.
Dates hide behind prices. Put a mark next to April fifteenth and September thirtieth as they pass, so a date-based question does not catch you having filed only the fees. One symbol is enough to make the date findable.
When the agent contrasts single tickets with a subscription, the contrast itself is testable. Write the thing that only one option gets, because Part 3 questions often reward the listener who tracked the comparison rather than the raw numbers.
Call a box office or read a rec centre's fees page aloud and quiz yourself on the details a minute later. Part 3 is the everyday act of getting information from someone whose job is to give it, and that is easy to practise in the wild.
When you reveal a clip, the teaching is in the notes that name the source of each wrong number: this option is the large plot's fee, this one is the per-play price. Reading where every distractor came from is how you learn to spot the borrowed figure before it borrows a mark from you.
Part 3 is a conversation between two people, one woman and one man, where one of them has information the other wants: a coordinator, a clerk, an agent explaining how something works. It runs about two to two and a half minutes and carries six questions. The clip is dense with concrete facts, so the work is holding on to the exact prices, dates and conditions.
Six, all from one clip. Each is a four-option multiple-choice question, and they usually follow the order in which the facts were mentioned. This course gives you two worked Information clips, a garden plot and a theatre subscription, each with its questions and a written reason for every option.
Because most of them are real numbers from the same clip. A typical Part 3 distractor is a genuine figure lifted from another moment in the talk: the larger plot's fee offered for the smaller plot, or the per-play price offered as the season price. The question is never whether a number was said, but which fact it belongs to.
On the real 2026 test the six questions are spoken after the clip, one at a time, with four written options and about thirty seconds each, answered in order. In this practice course the questions are printed so you can reread them, because there is no timer and the aim is to see how each figure was tied to the right fact.
No. Nothing here is timed and nothing is marked. You listen, settle on an answer, then reveal the transcript and the reasoning for all four options, including which mishearing each wrong number was built to catch. The timed listening tests are where your answers are scored and turned into a CLB band.
The scripts are Canadian, set in London, Ontario, at a community garden and a neighbourhood playhouse, with fees and dates in Canadian terms. The voices are North American, because no Canadian text-to-speech voice is available on any provider we can reach. Everything follows the 2026 CELPIP format.
Part 4 drops to a single voice and switches the questions to sentence completion, so the engine itself changes. Then meet the three-speaker discussion, and take the whole section under the real clock.
One news reader, facts in order, and questions you read and complete rather than hear.
Learn Part 4 → 2Two people who know each other, where the feeling sits under the arrangement.
Learn Part 2 → 5Three speakers who disagree, and questions that turn on who sided with whom.
Learn Part 5 →All 38 questions across the six parts, played once under the real conditions.
Start a listening test →Play one Information clip, keep the figures straight with no clock running, then reveal the transcript and the reason behind every option. Both clips are free.
