Listening Closely: A Guide to the C2 Proficiency (CPE)
Philo and Sophia break down the hardest listening paper in the Cambridge suite.
Approx. 40 minutes4 parts · 30 questionsEach recording heard twice
Part 1Multiple choice — three short unrelated extracts, 2 questions each (6 total)
Part 2Sentence completion — one long monologue, 9 questions
Part 3Multiple choice — one conversation between interacting speakers, 5 questions
Part 4Multiple matching — five short themed monologues, two tasks, 10 questions
Philo
Sophia, tell me honestly — is the CPE Listening paper actually harder than First and Advanced, or does it just feel that way because everything else about C2 is harder?
Sophia
Both are true. The shape is nearly identical to First and Advanced — four parts, the same task types you already know. But the speech is faster, the vocabulary is rarer, and the ideas being discussed are more abstract. You are no longer listening for "what happened." You are listening for attitude, implication, and the relationship between two speakers' opinions.
Philo
So the format is familiar, but the listening itself has to be sharper.
Sophia
Exactly. Let's walk through all four parts.
1 Multiple Choice: Three Short Extracts (Questions 1–6)
You hear three separate, unrelated extracts — monologues or exchanges between two speakers — each about a minute long, each with two three-option multiple-choice questions. A short context sentence introduces each extract, telling you the situation and who is speaking.
Context: You hear part of an interview with a woman who restores ancient manuscripts.
Interviewer: "What's the hardest part of the job — is it the physical restoration itself?" Restorer: "People assume so, but honestly, the physical work is almost meditative once you've trained your hands. The hardest part is the judgment calls — deciding how much of your own interpretation to allow into a damaged passage. You're not just repairing parchment. You're making decisions that will shape how future scholars read a text. That responsibility never gets lighter, no matter how many manuscripts you've handled."
What the question tests: A question here might ask what the restorer finds most difficult about her work. The trap answer is "the physical restoration" — mentioned but explicitly dismissed ("almost meditative"). The correct answer is the interpretive responsibility. In Part 1, the first idea mentioned is very often the distractor, not the answer.
Strategy: Because each extract is short and self-contained, use the gap between extracts to fully reset — clear the previous extract from your mind before the context sentence for the next one begins. Mixing information between extracts is a common source of lost marks.
2 Sentence Completion (Questions 7–15)
One monologue, three to four minutes long — a talk, a lecture, or a detailed personal account — with a longer gap-fill task: nine sentences with a word or short phrase missing, which you complete using the exact words you hear.
"...and what surprised me most, going back through the original excavation notes, wasn't the pottery or even the coins — it was the sheer volume of everyday correspondence we'd overlooked. Letters between merchants, complaints about shipments, a note from a mother to her son. We'd spent decades cataloguing the monumental inscriptions and almost missed the informal record sitting in the same boxes. It's the ordinary material that tells you how people actually lived, not just how they wanted to be remembered."
9. The excavation notes revealed a large quantity of overlooked . (Answer: correspondence / letters — captures what was "overlooked," not the pottery or coins, which are explicitly set aside)
Strategy: At C2, the exact wording you need is often introduced by a contrast signal — "but," "however," "what surprised me was," "we almost missed." These phrases flag that the upcoming words are the ones the question wants, distinguishing them from the surrounding detail that exists only to set the scene.
A conversation or discussion between two speakers, three to four minutes, with five four-option multiple-choice questions. The questions focus on the speakers' opinions, attitudes, feelings, and the relationship between what each speaker believes — including where they agree, disagree, or only partly agree.
The C2-specific challenge: tracking two independent viewpoints.
By C2, it's rarely enough to know what one speaker thinks. A question might ask: "The two speakers agree that... " or "Speaker B's attitude towards Speaker A's proposal is best described as..." — which requires holding both people's positions in mind simultaneously and noticing exactly where they diverge.
A useful habit: as you listen, silently note each speaker's stance in one or two words — "cautious," "enthusiastic but skeptical of cost." Small distinctions like "skeptical of cost" versus "skeptical overall" are exactly what separates the correct option from a close distractor.
4 Multiple Matching (Questions 21–30)
Five short, thematically linked monologues (about 30 seconds each), heard twice. Unusually, this part carries two separate matching tasks from the same five recordings — for example, matching each speaker to their opinion on a shared topic, and separately matching each speaker to a specific detail they mention. This gives ten questions from five recordings.
Strategy: Because there are two tasks running from the same audio, split your attention deliberately: on the first listening, focus mainly on Task One (options A–H for the first question set); on the second listening, shift your primary focus to Task Two, using your first-listening notes to fill any Task One gaps you're still missing. Trying to complete both tasks fully on every single listening usually means completing neither well.
A note on options: As with First and Advanced, there are always more options than speakers (typically eight options for five speakers), and in the second task, options may be used more than once or not at all. Never assume unused options in Task One are wrong for Task Two — they are independent tasks.
How Listening Is Scored
Every question is worth one mark, for a total of 30 marks, which is then scaled into the overall Listening component score. There's no separate mark for spelling in Part 2 beyond recognisable accuracy, but an answer must be unambiguous — if a spelling error creates a different existing English word, it will be marked wrong.
6Part 1
9Part 2
5Part 3
10Part 4
Sophia
One last thought, Philo. At every level below C2, listening tests whether you can follow English. At C2, it tests whether you can follow people — their attitudes, their disagreements, the things they imply rather than state outright.
Philo
Which is, I suppose, the harder skill in any language — including our own.