What Part 2 asks you to do, in one paragraph
In Part 2 of the B2 First and C1 Advanced Speaking tests you talk on your own for 1 minute, then answer a short question about your partner's photos for about 30 seconds. At B2 First you get two photographs with one question printed above them, and the B2 First handbook says candidates "are expected to point out similarities and differences between the photographs and then move on to deal with the question, answering it with reference to both photographs". At C1 Advanced you get three pictures, choose two, and answer two printed questions; the C1 Advanced handbook says candidates "will always be asked to speculate" and "will never be asked to merely describe the visual". So the phrases that earn marks are comparing and speculating phrases, not describing ones. This page gives you 40 of them grouped by job, a timeline for the 60 seconds, two complete long turns written by TalkDrill (one B2, one C1), the upgrades that move a C1 answer towards 200, and what Cambridge's own examiner comments say about real Part 2 performances.
- B2 First: two colour photographs, one printed question, 1 minute each, then about 30 seconds on a question about your partner's photos. Part 2 lasts 4 minutes for a pair.
- C1 Advanced: three pictures, talk about any two, two printed questions, 1 minute, then about 30 seconds on your partner's set. 4 minutes for a pair, 6 for a group of three.
- B1 Preliminary is a different task: one photograph, describe it for about 1 minute, no comparing and no speculating.
- Do not speak during your partner's minute. Do ask the interlocutor to repeat the instructions if you need to: the B2 handbook says you "will not lose marks" for it.
- Keep talking until the interlocutor says "Thank you". The C1 handbook warns against "closure techniques such as 'That's all!'".
Part 2 at B1, B2 and C1, side by side
Part 2 changes shape between the three exams more than any other part, which matters if you passed B1 Preliminary and are now preparing for B2 First, or are deciding between B2 and C1. The rows below come from the exam-format pages and the handbooks.
| Feature | B1 Preliminary | B2 First | C1 Advanced |
|---|---|---|---|
| What you are given | One colour photograph | Two colour photographs | Three pictures; you choose two |
| What you must do | Describe what you can see | Compare the photos, then answer the printed question about both | Compare two pictures, then answer both printed questions about both |
| Printed above the pictures | Nothing | One direct question | Two questions |
| Speculate? | No: "should not speculate about the context" | Yes, to answer the question | Always: "will never be asked to merely describe" |
| Your time | About 1 minute | 1 minute | 1 minute |
| Your partner's turn | Listens; no response task | About 30 seconds on a short question about your photos | About 30 seconds on a short question about your pictures |
| If you stop early | The interlocutor has back-up prompts: "Talk about the people", "the place", "other things" | You give the examiners less to mark: the handbook warns of failing "to give the examiners an adequate sample of language" | Keep going: the handbook says keep talking "until the interlocutor says 'Thank you'" |
| Whole part | 3 minutes for the pair | 4 minutes for the pair | 4 minutes (6 for three) |
The B2 handbook describes the focus as "organising a larger unit of discourse; comparing, describing and expressing opinions", and the C1 handbook adds "speculating". The B1 handbook says the opposite about description: describe "as fully as possible", as if to someone who cannot see the photograph, including "colours, clothing, time of day, weather". In our judgement, the habit that most often follows a candidate from B1 into B2 is that full description: forty seconds of "there is a man, he is wearing a blue shirt" before the question appears. The B1 Preliminary Speaking page covers the B1 task; everything below is about B2 and C1.
The 1-minute structure: six moves
A minute holds about six short moves. The timings are our recommendation, not a Cambridge rule, but they follow the handbook's order: similarities and differences first, then the question, answered "with reference to both photographs".
| Seconds | Move | What it sounds like (B2) | At C1 |
|---|---|---|---|
| 0 to 5 | 1. Frame | "Both photos show people ..., but in very different situations." | Say which two you have chosen and why they make a good pair. |
| 5 to 15 | 2. One similarity | "One similarity is that ..." | Fold moves 2 and 3 into one sentence with "whereas" or "by contrast". |
| 15 to 25 | 3. One or two differences | "The main difference is that ..., whereas ..." | Pick the difference that the first question will use. |
| 25 to 40 | 4. The question, photo A | "As for why they ..., I'd imagine ..." | Question 1 for both pictures. |
| 40 to 52 | 5. The question, photo B | "The woman in the second, on the other hand, might ..." | Question 2 for both pictures. |
| 52 to 60 | 6. Round off and keep going | "So in both cases ..., but for different reasons." | A concession: "Although, to be fair, ..." |
Why 25 seconds? Because the printed question is where the content of your minute lives. Describing past that point spends the time the handbook says should go on dealing with the question. If your recordings show you reaching the question at 40 seconds, cut move 2 to a single clause: "Both photos show people eating together, but ...".
The 40 phrases, grouped by job
Phrases 6 and 11 are the B2 First handbook's own examples of language for similarity and difference. The other 38 are TalkDrill's selection. Do not learn all forty: pick two per group that feel natural to you and use them until they are automatic. The C1 handbook suggests students "build up their own lists of suitable phrases throughout the course", and the B2 handbook ties this language to marks directly: "using comparatives and linking words will produce a more extended and coherent sample of speech than simply stringing together a series of simple statements. This will help them to gain marks under the Discourse Management assessment criterion."
Framing the comparison, and choosing at C1 (1 to 5)
| # | Phrase | Example (TalkDrill) | Level |
|---|---|---|---|
| 1 | Both photos show ..., but in very different situations. | Both photos show people exercising, but in very different situations. | B2 |
| 2 | What these two photos have in common is ... | What these two photos have in common is that everyone looks focused. | B2 |
| 3 | At first sight they look quite similar, but ... | At first sight they look quite similar, but the mood is completely different. | B2 |
| 4 | I'll talk about the first and the third pictures, because ... | I'll talk about the first and the third pictures, because they show opposite kinds of pressure. | C1 |
| 5 | These two seem to me to show opposite ways of ... | These two seem to me to show opposite ways of spending a Saturday. | C1 |
Similarities (6 to 10)
| # | Phrase | Example (TalkDrill) | Level |
|---|---|---|---|
| 6 | One similarity is that ... (handbook) | One similarity is that both groups are outdoors. | B2 |
| 7 | In both cases, ... | In both cases, the people seem to have chosen this freely. | B2 |
| 8 | Neither of them seems to ... | Neither of them seems to be in a hurry. | B2 |
| 9 | Just like ..., the ... | Just like the family, the two friends are sharing their food. | B2 |
| 10 | They're alike in the sense that ... | They're alike in the sense that both jobs need a lot of patience. | C1 |
Differences (11 to 17)
| # | Phrase | Example (TalkDrill) | Level |
|---|---|---|---|
| 11 | In this picture there's ..., whereas in the other there's ... (handbook) | In this picture there's a big crowd, whereas in the other there's just one runner. | B2 |
| 12 | The main difference is that ... | The main difference is that one is planned and the other isn't. | B2 |
| 13 | In the first photo ..., while in the second ... | In the first photo they're indoors, while in the second she's by the sea. | B2 |
| 14 | ..., on the other hand, ... | The class, on the other hand, looks noisy and sociable. | B2 |
| 15 | Unlike ..., ... | Unlike the lecture, the violin lesson is one to one. | B2 |
| 16 | ... is far more ... than ... | The street stall is far more relaxed than the family lunch. | B2 |
| 17 | By contrast, ... | By contrast, the controller works completely alone. | C1 |
Speculating (18 to 26)
| # | Phrase | Example (TalkDrill) | Level |
|---|---|---|---|
| 18 | They might be ... / They may have ... | They might be training for a race. | B2 |
| 19 | They must be ... (when you are almost sure) | She must be an early riser. | B2 |
| 20 | They can't be ... (when you are almost sure it is not true) | They can't be beginners; they all know the steps. | B2 |
| 21 | It looks as if ... / It looks as though ... | It looks as if the music is very loud. | B2 |
| 22 | Judging by ..., I'd say ... | Judging by her headphones, I'd say she wants to be alone. | B2 |
| 23 | I'd imagine that ... | I'd imagine that the class is straight after work. | B2 |
| 24 | I get the impression that ... | I get the impression that he does this every single day. | C1 |
| 25 | Presumably, ... | Presumably, they've been queuing since before dawn. | C1 |
| 26 | They may well have ... | She may well have chosen the beach to avoid people. | C1 |
Answering the printed question (27 to 32)
| # | Phrase | Example (TalkDrill) | Level |
|---|---|---|---|
| 27 | As for why they ..., I think ... | As for why they chose this, I think it's the company. | B2 |
| 28 | Turning to the question, ... | Turning to the question, the people in the class are probably enjoying the energy. | B2 |
| 29 | That's probably why ... | That's probably why she's out so early. | B2 |
| 30 | In both cases, I suspect the main reason is ... | In both cases, I suspect the main reason is that they simply enjoy it. | B2 |
| 31 | Another thing that might matter is ... | Another thing that might matter is the cost. | B2 |
| 32 | If I had to choose, I'd say ... | If I had to choose, I'd say the runner is getting more out of it. | B2 |
Rounding off and repairing (33 to 36)
| # | Phrase | Example (TalkDrill) | Level |
|---|---|---|---|
| 33 | So in both cases ..., but ... | So in both cases they're enjoying it, but for opposite reasons. | B2 |
| 34 | What I mean is ... | What I mean is, she doesn't need anyone else to keep going. | B2 |
| 35 | I'm not sure what it's called, but it's a kind of ... | I'm not sure what it's called, but it's a kind of raised platform at the front. | B2 |
| 36 | Sorry, could you repeat the question, please? | Use it before you start, not halfway through. | B2 |
Phrase 35 matters more than it looks. The B2 handbook tells teachers to "train your students to paraphrase when they do not know, or cannot remember, a word", and the B1 handbook says candidates "will be given credit" for doing so. Freezing on a missing noun costs more than describing it.
The 30-second answer about your partner's photos (37 to 40)
| # | Phrase | Example (TalkDrill) | Level |
|---|---|---|---|
| 37 | I'd say ..., mainly because ... | I'd say the beach, mainly because I can't stand crowds. | B2 |
| 38 | Personally, I'd prefer ..., although ... | Personally, I'd prefer the class, although I'd feel shy at first. | B2 |
| 39 | It's hard to choose, but if I had to, ... | It's hard to choose, but if I had to, I'd go for the street food. | B2 |
| 40 | For me it would be ..., because when I ... | For me it would be the class, because when I run alone I give up after ten minutes. | B2 |
Thirty seconds is three or four sentences: an answer, a reason, a small example or a "having said that". "The first one" and silence gives the assessor nothing to mark.
The describe-versus-compare trap
Here is a weak opening to the photos in worked long turn 1 below, written by TalkDrill to show the pattern. It is accurate English, and it is not what the task asks for.
Weak: 25 seconds of description (TalkDrill)
"In the first picture there are a lot of people in a gym. They are wearing sports clothes. There is a woman at the front, she is the teacher. There are mirrors on the wall. In the second picture there is a woman. She is running on the beach. She has headphones. The sky is orange. The sea is calm."
Nine sentences, no comparison, no link between them, and the printed question ("What might the people be enjoying about exercising in these ways?") has not been touched. The B2 handbook is explicit: candidates "are not expected to give detailed descriptions of each picture. Rather, they are asked to compare the pictures and to give their reaction to them." The same information, used to compare and to reach the question:
Strong: the same facts in 12 seconds (TalkDrill)
"Both photos show people exercising, but in very different ways: a big, noisy class in a gym, whereas in the second there's one woman running alone on an empty beach at sunrise. As for what they might be enjoying, ..."
At C1 the gap is larger. The C1 handbook gives its own contrast: "This picture shows two people who are on an expedition. They have to work together to achieve their goal." is "not as impressive as" a version that begins "These people look like they're on an expedition to the North Pole or they might be climbing a mountain together", goes on to guess at the reasons and the relationship, and ends "Their lives could depend on how well they co-operate with each other if there is an emergency." Every sentence of the stronger version is a guess with a reason.
Worked long turn 1: B2 First (TalkDrill)
Photo 1: an indoor fitness class of about twenty people following an instructor on a small stage, mirrors along the wall, speakers in the corners. Photo 2: a woman running alone along an empty beach at sunrise, headphones on, the tide out.
Printed question: What might the people be enjoying about exercising in these ways?
Question for the partner afterwards: Which way of exercising would you prefer? (Why?)
The long turn, about 165 words (TalkDrill model, not a real candidate)
[0:00] Both photos show people exercising, but in very different situations. [0:05] One similarity is that everyone seems to be taking it seriously; nobody is just chatting. [0:12] The main difference, though, is that the first is a big indoor class, with about twenty people following an instructor to loud music, whereas in the second there's just one woman running along an empty beach, very early in the morning.
[0:24] As for what they might be enjoying, I'd imagine the people in the class enjoy the energy of the group. When the music is loud and everyone is moving together, it's probably easier to keep going than on your own, and some of them may have come with friends. [0:40] The runner, on the other hand, seems to be enjoying the exact opposite: the silence and having time to herself. Judging by the sunrise, she must go out before work, so it might be the only quiet moment in her day.
[0:53] So in both cases they're enjoying the exercise, but for almost opposite reasons: one is about people, and the other is about getting away from them.
| Moment | What the assessor hears | Criterion it feeds |
|---|---|---|
| "Both photos show ..., but ..." | The task understood in five seconds | Discourse Management: relevance and organisation |
| "whereas in the second there's just one woman" | A comparison, not a list | Discourse Management: cohesive devices |
| "As for what they might be enjoying" | The printed question reached at 24 seconds | Discourse Management: "contributions are relevant" |
| "it's probably easier to keep going than on your own" | A comparative inside a reason | Grammar and Vocabulary |
| "she must go out before work, so it might be" | Two degrees of certainty in one sentence | Grammar and Vocabulary: some complex forms |
| "one is about people, and the other is about getting away from them" | An ending that answers the question for both photos | Discourse Management; stress on PEOPLE and AWAY for Pronunciation |
The partner's 30 seconds (TalkDrill model)
Interlocutor: "Which way of exercising would you prefer?" Candidate B: "I'd go for the class, mainly because I'm not very disciplined. If I go running on my own, I usually find an excuse to stop after ten minutes, but with an instructor and other people around me, I'd feel embarrassed to give up. Having said that, I can see why the beach would be more relaxing."
Worked long turn 2: C1 Advanced (TalkDrill)
Picture 1: a teenage girl giving a speech at a school assembly, hundreds of students on benches, notes in her hand. Picture 2: a street musician playing guitar to a small crowd in a railway station, his guitar case open on the ground with some coins in it. Picture 3: a chef presenting a plate to three judges in a television kitchen, cameras around them.
Printed questions: Why might the people have decided to perform in front of others? How might they be feeling?
Question for the partner afterwards: Which of these people do you think has the most to lose? (Why?)
The long turn, about 200 words (TalkDrill model, not a real candidate)
I'll talk about the girl giving the speech and the street musician, because they seem to me to be performing for completely different reasons. In both cases there's an audience, but the girl's audience has no choice about being there, whereas the people at the station have stopped voluntarily and could walk away at any moment.
As for why they decided to perform, the student may well not have decided at all; she might have been chosen by a teacher, or she could be standing for the school council and need the votes. The musician, by contrast, has presumably chosen this deliberately, either to earn some money, judging by the open guitar case, or simply because he loves playing and wants people to hear him.
How might they be feeling? I'd imagine the girl is fairly nervous. She's clutching her notes and she's in front of hundreds of people who know her, which in a way is harder than performing for strangers. The musician looks far more relaxed, although I get the impression he's keeping an eye on the crowd to see whether anyone will stop. So she's probably the more anxious of the two, but he's the one who depends on the audience.
| Moment | Why it is C1 rather than B2 (our reading of the descriptors) |
|---|---|
| "because they seem to me to be performing for completely different reasons" | The choice of pictures is justified, so the frame already answers part of question 1. |
| "may well not have decided at all" | A modal with "well" and a negative that challenges the question's assumption: range and flexibility. |
| "she could be standing for the school council" | A specific, plausible guess instead of "maybe she wants to". |
| "by contrast", "presumably", "judging by" | Three different cohesive devices and stance markers in one sentence. |
| "clutching her notes" | A precise verb: Lexical Resource, which C1 marks separately. |
| "which in a way is harder than performing for strangers" | A non-obvious idea, qualified ("in a way"). |
| "she's probably the more anxious of the two, but he's the one who depends on the audience" | A closing contrast that answers both questions for both pictures. |
The partner's 30 seconds (TalkDrill model)
Interlocutor: "Which of these people do you think has the most to lose?" Candidate B: "I'd say the chef, actually, the one my partner didn't choose. A bad school speech is forgotten by Monday, and a busker can move to another station, but if you fail in front of television judges it's recorded, and it could follow you around for years. That said, the girl probably feels as though she has the most to lose."
Notice that the partner's answer uses the third picture. The C1 interlocutor's question is about the set, so listen to your partner's long turn with all three pictures in mind.
C1 upgrades: from a 180 answer towards 200
The C1 Advanced handbook puts Grades B and C at 180 to 199 and Grade A at 200 to 210. Cambridge does not publish phrase lists for each score; the table below is our reading of the C1 Band 5 descriptors, which ask for control of "a wide range of grammatical forms", "a wide range of appropriate vocabulary to give and exchange views on familiar and unfamiliar topics", "a wide range of cohesive devices and discourse markers", contributions that are "relevant, coherent and varied", and "phonological features" used "effectively to convey and enhance meaning".
| Typical B2-level line | C1 upgrade | What changes |
|---|---|---|
| Maybe they are tired. | They may well be exhausted after a long shift. | Graded certainty, a precise adjective, a reason built in. |
| I think ... | I'd imagine ... / My guess would be ... / I suspect ... | Variety of stance markers instead of one repeated. |
| ..., but ... | ..., whereas ... / By contrast, ... / That said, ... | A wider range of cohesive devices. |
| The photos are very different. | The two situations could hardly be more different. | A complex form that carries emphasis. |
| They are happy. | They seem to be thoroughly enjoying themselves. | Collocation and a softened claim. |
| In the first photo there is ... | What strikes me first is ... | A cleft structure that organises the turn. |
| It is difficult. | It can't be easy to keep that up all day. | Speculation through a negative modal. |
| That's all. | Although, to be fair, ... | A concession that uses the last seconds instead of closing early. |
Use one upgrade per move, not one per sentence. A minute in which every clause is decorated sounds rehearsed, and the C1 handbook asks for language that is "relevant, coherent and varied", not dense. The C1 Advanced Speaking page sets a 180 and a 200 answer side by side for every part.
What Cambridge examiners said about real Part 2 turns
Cambridge publishes Speaking test videos with examiner comments on its preparation pages. For B2 First the videos are of Giulia, Tabatha and Elena (a group of three), Julia and Isabela, and Natalia and Enzo; for C1 Advanced, Marta, Giovanna and Anna, Pedro and Isabela, and Teresa and Antonio. We read the examiner comments for Giulia, Tabatha and Elena (Cambridge PDF, dated 2024, linked from the B2 First preparation page, read 21 October 2026). Four of its observations, three of them about Part 2 turns, are worth more than any tips list, because they show what a real assessor notices in a minute.
| Candidate and score | What the examiners wrote (Part 2 unless marked) | The drill it suggests |
|---|---|---|
| Giulia, Pronunciation 4.5 | "Occasionally, she uses rising intonation inappropriately at the end of statements", with the Part 2 example "I think they're in Japan." | Record your long turn and check that statements fall at the end. A rising tone makes a speculation sound like a question to the examiner. |
| Giulia, Discourse Management 4 | Across the whole test, not only Part 2: uses "a range of appropriate cohesive devices" but "there is overuse of fillers, such as 'like' and 'you know'". | Swap the filler for a planned move phrase from the tables above. |
| Tabatha, Grammar and Vocabulary 3.5 | "She attempts a few complex forms, but sometimes with limited success", for example "more enjoyable to visit the city on (from a) different point of view (part 2)". | Attempting complex forms is in the B2 Band 3 descriptor; keep attempting, and drill the prepositions in your favourite comparison phrases. |
| Elena, Discourse Management 4 | She "organises extended discourse easily, despite faltering a little in part 2", and one Part 2 contribution lacked coherence: "...to improve the matches or their entertain, I don't know." | When a sentence collapses, repair it with phrase 34 ("What I mean is ...") rather than "I don't know". |
All three scored 3.5 to 4.5 on every criterion in that commentary, so these are the remarks examiners make about good B2 candidates, not weak ones. Watch the video on the preparation page with the comments open; it is the closest you will get to sitting beside an assessor.
First-language traps in the long turn
These are TalkDrill's judgement, drawn from our research on the first languages of Cambridge candidates, not from Cambridge data. The B2 Band 5 Pronunciation descriptor asks for intelligibility, appropriate intonation, accurately placed stress and clearly articulated sounds; it says nothing about accent.
| First language | What tends to slip in a 1-minute turn | Fix |
|---|---|---|
| Spanish and Catalan | "The people is", flat intonation across a long turn, "in the photo of the left" | "People are"; "on the left"; let the voice fall at each full stop |
| Italian | Extra vowels after final consonants ("photo-a"), "I am agree", fast delivery that blurs the comparison | Slow down on the contrast words; "I agree" |
| Portuguese | "In the first photo have a family" for "there is", final /l/ sounding like /w/ | "In the first photo there's a family" |
| Turkish | The main point arriving late, after the reasons; missing articles ("in first photo") | Say the answer first, then the reason; "in the first photo" |
| Vietnamese | Dropped final consonants, so plurals and past tenses disappear ("both photo"); turns that stop at 40 seconds | Exaggerate final /s/ and /t/ in practice; use the six-move plan to fill the minute |
| Greek | Long and short vowels merging (ship and sheep, full and fool), because Greek has five vowels; stress moved on borrowed words | Practise minimal pairs from your own photo descriptions; check stress in a dictionary for words such as "photograph" and "photographer" |
| Polish | Articles ("in second picture"), /th/ in "both", "they", "the other" | Drill phrases 1, 6, 11 and 13 slowly until the articles and /th/ are automatic |
The fuller lists, with drills, are on the pages for Spanish, Italian and Portuguese speakers and Turkish and Vietnamese speakers.
A seven-day Part 2 practice route
Fifteen to twenty minutes a day. Record everything on your phone; the recording is the teacher.
- Day 1, measure. Run bank task B2-03 (or C1-01 at C1) with a timer. Note the second at which you first answer the printed question. That number is your baseline.
- Day 2, phrases. Choose two phrases from each group above. Say each one aloud five times with the stress on the contrast word, then use all of them on one photo pair from the official sample paper.
- Day 3, the six moves. Run B2-04 or C1-02, pausing the timer after each move to check you are on schedule.
- Day 4, with a partner. Do a full Part 2 for both candidates (4 minutes) with a friend, or in TalkDrill's role-play practice with the AI as the interlocutor asking your 30-second question.
- Day 5, sound. Listen to your day 4 recording for intonation only, as the examiners did for Giulia. Then read a paragraph aloud in read-aloud practice, which scores pronunciation, and focus on falling tones at full stops.
- Day 6, the official video. Watch one Speaking test video on the preparation page and read its examiner comments.
- Day 7, re-measure. Repeat day 1's task. If you reach the question earlier than on day 1, the plan is working.
Then move on to Part 3, which is where Interactive Communication is won. The full set of 30 timed tasks is in the practice bank, and every Cambridge page is linked from the Cambridge English hub. If you are sitting a paper-based B2 First on 7 November, 24 November or 12 December 2026, or C1 Advanced on one of its November and December dates, ask your centre now for your Speaking date, which can be on a different day from the written papers: the official dates page says Speaking test dates "are subject to local provision". Official sources for this page: the B2 First exam-format page, the C1 Advanced exam-format page and the B2 First handbook for teachers.