TOEFL 2026 Speaking Section: Complete Guide

TTtoeflprep.ai Teamon February 17, 202628 min read
TOEFL 2026 Speaking Section: Complete Guide

How the new TOEFL 2026 Speaking section works — two task types, exact format details, real sample questions with answers, scoring criteria, and a prep plan.

The TOEFL 2026 Speaking section is completely different from the old format.

The four old tasks — one independent and three integrated — are gone. In their place: two new tasks, 11 total items, and about 8 minutes of speaking time. No preparation time for anything. No reading or listening input to summarize. Just you, a microphone, and either a sentence to repeat or a question to answer immediately.

This guide covers both tasks in detail — exact format, real examples with sample answers, what scores are based on, and how to prepare.

Key facts at a glance

Duration: ~8 minutes · Items: 11 total · Tasks: 2 (Listen and Repeat + Take an Interview) · No prep time for either task · Not adaptive · Speaking is the last section of the test

What Changed

The Speaking section has been completely revised. All four old tasks have been replaced with two new tasks: Listen and Repeat and Take an Interview. This section now focuses heavily on spontaneous and real-time verbal communication.

FeatureOld TOEFL (2005–2025)TOEFL 2026
Total tasks4 (1 independent + 3 integrated)2
Total items411
Duration~16–17 minutes~8 minutes
Preparation time15–30 seconds per taskNone
Reading/Listening inputYes (integrated tasks)No
Task typesIndependent + integrated summariesListen and Repeat + Take an Interview
Section orderLastStill last
AdaptiveNoNo (only Reading and Listening are)
Scored byHuman ratersAI engine + human oversight

Speaking is NOT adaptive

Unlike Reading and Listening, the Speaking section does not change based on your performance. Everyone gets the same two tasks in the same format. There is no routing module, no hard path or easy path — just 11 items, same for everyone.

How Speaking Is Scored

The Speaking and Writing responses are scored by ETS's proprietary AI scoring engine according to the criteria outlined in the scoring guides. These engines use advanced natural language processing, and human rating remains a critical component — automated engines are trained on human ratings, and humans provide oversight.

ETS updated the scoring rubric to a 5-point scale for the new task types. The fundamentals are the same — delivery, language use, and topic development — but the criteria now reflect conversational fluency rather than structured responses.

The two tasks are scored differently:

Listen and Repeat: A perfect score requires that "the response is fully intelligible and an exact repetition of the prompt." Content doesn't matter here — only how accurately and clearly you reproduce the sounds.

Take an Interview: Scored on fluency, coherence, vocabulary range, grammar accuracy, and how well you develop and support your answers.

BandCEFR LevelWhat it means for Speaking
6C2Mastery — clear, fluent, accurate across both tasks
5–5.5C1Advanced — natural fluency, strong pronunciation
4–4.5B2Upper-intermediate — generally clear with some errors
3–3.5B1Intermediate — understandable but frequent errors or pauses
Below 3A2Basic — most programs won't accept this

Task 1: Listen and Repeat (Questions 1–7)

Exact format:

  • You hear seven sentences related to a campus or daily life situation
  • After each sentence there is a short pause — you must repeat exactly what you heard
  • Sentences get longer and more difficult as they go on — the first might be 4 seconds long, the last might be 6 seconds
  • You have 8–12 seconds of recording time per sentence
  • No preparation time between sentences
  • An image on screen illustrates the situation — arrows, highlights, or shapes may appear on the image to guide context
  • The sentences are audio only — not shown on screen as text

Typical scenarios:

  • Giving a tour of a campus location (art gallery, library, chemistry lab)
  • Explaining a procedure at a campus facility
  • Describing steps in a campus or daily life process

All seven sentences in one set are about the same scenario and image.

The Listen and Repeat task tests the ability to accurately reproduce English sounds and rhythms with minimal thought, relying on short-term memory.

Specifically:

  • Pronunciation accuracy — individual sounds, especially ones that differ from your native language
  • Word stress — stressing the right syllable in multi-syllable words
  • Sentence rhythm and intonation — the natural rise and fall of spoken English
  • Auditory memory — holding a 4–6 second sentence in your head long enough to repeat it accurately

What this does NOT test

Content, ideas, grammar, or vocabulary choices. You're not being asked to create anything — just to reproduce what you heard as accurately as possible. A sentence repeated perfectly but in a heavy accent can still score well if it's fully intelligible.

Scenario: Giving a tour of a local art gallery

You are learning how to welcome visitors to a local art gallery. Listen to the speaker and repeat what they say. Repeat only once.

The seven sentences, in order (increasing length and complexity):

  1. Welcome to the art gallery.
  2. A free audio guide is available for all visitors.
  3. Digital maps can be used for planning your visit.
  4. If you have questions, just ask a staff member.
  5. When taking photos, please turn off your flash.
  6. There's also a quiet area over here for personal reflection.
  7. Before leaving the gallery, please make sure to return your audio guide at the entrance.

Notice how sentence 1 is simple and short. By sentence 7 it's a full complex sentence with a subordinate clause. That progression happens in every Listen and Repeat set.

Before test day:

  • Practice recording yourself repeating small pieces of audio and listen back to check accuracy. Most people don't know what their spoken English actually sounds like — recording removes that blind spot.
  • Identify the specific sounds in English that are hard for you and focus on them directly. Common problem areas: th/d confusion, r/l distinction, vowel length, consonant clusters at the end of words.
  • Shadowing exercises are highly effective — listen to a sentence and speak at the same time as the audio, matching rhythm and intonation exactly. Do this daily.
  • Practice with sentences of increasing length. Start with 4-word sentences, work up to 12–15 word sentences.

On test day:

  • You have 8–12 seconds to record. You don't need to rush — wait a moment to gather yourself before speaking.
  • If you know you made a mistake, it's okay to self-correct within your recording time.
  • If you completely forget a word, make your best guess and keep going. Don't go silent — a partial attempt scores better than silence.
  • Focus on clarity over speed. Speak at a natural pace, not rushed.

Task 2: Take an Interview (Questions 8–11)

Exact format:

  • You participate in a simulated interview — you see a video of an interviewer who asks four questions about a topic connected to everyday life
  • You have 45 seconds to respond to each question
  • No preparation time — you must answer immediately after the question is asked
  • You can only hear the questions — you cannot read them on screen
  • The interviewer asks follow-up questions — questions get slightly harder as they go

Topics are always familiar:

Smartphones, study habits, transportation, campus life, hobbies, work preferences, technology, social habits. Nothing obscure or technical. The topic is introduced with a brief scenario at the start — for example: "You have agreed to take part in a research study about smartphone usage."

This section evaluates your ability to speak fluently and clearly, develop and expand on your ideas, and use accurate grammar and vocabulary in a conversational setting.

The four things that affect your score:

  • Fluency and pace — Can you speak for 45 seconds without long pauses or filler sounds?
  • Coherence — Does your answer make sense and stay on topic?
  • Grammar and vocabulary — Are your sentences grammatically correct? Do you use a range of words rather than repeating the same ones?
  • Development — Do you give reasons and examples, not just a one-sentence answer?

What graders look for vs. what students think they look for

Students often think complex vocabulary impresses graders. It doesn't — clarity and coherence do. A simple, well-developed answer with correct grammar beats a complicated, unclear one every time. Use words you're confident with.

The questions follow a consistent pattern:

QuestionTypeWhat it asks for
Q8Personal recollectionDescribe a specific memory or past experience related to the topic
Q9Personal preferenceExplain how you personally feel about or relate to the topic
Q10Opinion / stanceGive your view on whether something is good, bad, true, or false
Q11Policy / perspectiveShould institutions, governments, or companies do something? Why?

Each question builds on the previous one — they all stay on the same topic. By Q11 you're giving a broader opinion or recommendation about the topic you've been discussing since Q8.

Useful opening phrases by question type:

QuestionUseful openers
Q8 (recollection)"A few weeks ago..." / "I remember when..." / "Last year, I..."
Q9 (preference)"Well, for me..." / "Personally, I tend to..." / "I guess that's because..."
Q10 (stance)"Personally, I think that..." / "While some people believe X, I think..."
Q11 (policy)"Yes, they probably should..." / "No, I don't think that's a good idea because..." / "While it could be argued that X, I feel..."

Topic: Smartphone usage

You have agreed to take part in a research study about smartphone usage. The researcher will ask you some questions.

Q8 — Personal recollection:

"Think back to the last time you used your phone for something important. Why did you use it? What did you like about how it worked?"

Sample answer:

A few weeks ago I was traveling in a completely unfamiliar city — Paris, actually — and I got lost. I couldn't figure out how to get back to my hotel and I started feeling quite anxious. Fortunately I had my phone with me, so I opened a map app, entered my hotel's address, and got instant walking directions. It even showed me public transport options. The whole thing took about 30 seconds. That made me feel a lot better — what could have been a stressful situation turned into a minor inconvenience.

Q9 — Personal preference:

"Some people feel phones make them more connected and efficient. Others feel distracted or overwhelmed. How do you usually react to your phone in daily life?"

Sample answer:

Honestly, my phone often makes me feel more distracted than focused. I carry it everywhere and find myself checking it constantly — emails, messages, news — even when I should be concentrating on something else. I know it's a problem. When I'm somewhere beautiful, like a park or a museum, I still feel the urge to check my phone instead of just being present. It's something I'm aware of but haven't fully solved yet.

Q10 — Opinion / stance:

"Some people believe smartphones clearly make life better in the modern world. Do you agree? Why or why not?"

Sample answer:

Overall, yes — I think smartphones make life better, even though they cause distraction like I mentioned. The main reason is access to information. A student in a remote area with limited library resources can access the same academic materials as someone at a top university, just through their phone. That kind of access didn't exist 20 years ago. The distraction issue is real, but it's a problem with how we use phones, not with the technology itself.

Q11 — Policy / perspective:

"Should schools and workplaces encourage healthier phone habits — for example, asking people to turn off their phones during breaks? Why or why not?"

Sample answer:

I think that's a reasonable idea. Many people — myself included — experience anxiety from the constant flow of information coming from their phones. Encouraging some phone-free time during the day won't solve major issues, but it could meaningfully reduce stress on a small scale. It might also help students focus better on their work and help employees be more productive. As long as it's a suggestion and not a strict rule, I think most people would respond well to it.

Structure for 45 seconds:

45 seconds is roughly 100–120 words of natural speech. That's enough for:

  • One clear position or statement (1–2 sentences)
  • One specific reason (2–3 sentences)
  • One concrete example or detail (2–3 sentences)
  • A brief closing (1 sentence, optional)

Don't try to cover three reasons. One reason with a real example is stronger and more coherent than three vague points rushed in 45 seconds.

What to avoid:

  • Don't include long lists of things — graders don't respond well to that.
  • Don't memorize full scripts — they sound robotic and fall apart when you're nervous.
  • Don't repeat the same words more than twice — vary your vocabulary even slightly.
  • Don't go silent mid-answer. If you lose your train of thought, keep speaking — say "let me put it this way..." or "what I mean is..." and continue.

Pace and delivery:

  • Maintain a natural speaking pace — not too slow, not too fast.
  • Avoid long pauses. Speak without major interruptions.
  • Use basic transitions to give your answer structure: "moreover," "as a result," "for example," "on the other hand."
  • Use a variety of grammatical forms — include some subordinating conjunctions (because, although, which) to add complexity without overcomplicating.

Common Mistakes That Cost Points

Week-by-Week Practice Plan

Week 1 — Build the fundamentals separately

For Listen and Repeat: start shadowing daily. Find a podcast or YouTube channel with clear, natural English speech. Listen to a sentence, pause, repeat it out loud. Record yourself and compare. Focus specifically on sounds that are difficult for your native language background.

For Take an Interview: practice answering one question out loud every day for 45 seconds. Start with Q8-style questions (tell me about a time when...). Don't write answers — speak them directly. Record and listen back.

Week 2 — Add full sets under time pressure

Practice full Listen and Repeat sets — all 7 sentences in order, increasing in length. Time yourself. Practice full Take an Interview sets — all 4 questions on one topic, back to back with no preparation.

Identify your weakest area: Is it pronunciation? Running out of things to say? Long pauses? The answer changes what you practice most in weeks 3–4.

Week 3 — Targeted improvement

Focus 70% of your Speaking practice time on whichever task is weaker.

If Listen and Repeat is the issue: drill specific sounds. Use minimal pairs practice (ship/sheep, bed/bad, think/sink). Do 20 minutes of shadowing per day.

If Take an Interview is the issue: practice the 4-question pattern on 2–3 new topics per day. Time each answer. Review for coherence — did you stay on topic, give a real example, use varied vocabulary?

Week 4 — Full test simulation

Do full test simulations — Reading, Listening, Writing, then Speaking last. You need to practice Speaking when you're already mentally tired from the other sections. That's what test day actually feels like.

In your final week, stop learning new content. Consolidate what works. Practice the pacing, the transitions, and the confidence to keep speaking even when you're uncertain.

Old vs New Speaking Section

FeatureOld TOEFL (2005–2025)TOEFL 2026
Number of tasks42
Total items411
Duration~16–17 minutes~8 minutes
Prep time15–30 seconds per taskNone
Integrated inputReading + Listening before speakingNo input — speak immediately
Task typesIndependent + integrated summariesListen and Repeat + Take an Interview
Pronunciation taskNoYes (Listen and Repeat)
AdaptiveNoNo
AI scoringPartialPrimary (with human oversight)
Section positionLastStill last

Common Questions

Where to Focus Your Prep

Priority 1: Record yourself daily

You can't improve what you can't hear. Record every practice response — both Listen and Repeat and Take an Interview. Listen back and identify specific issues before your next session.

Priority 2: Fix your specific pronunciation gaps

Identify 3–5 sounds in English that you consistently get wrong. Practice those specifically. Shadowing is the most effective method — match the audio exactly while it plays.

Priority 3: Practice speaking for the full 45 seconds

Most students underuse their time in Take an Interview. One reason with one real example, spoken clearly for 45 seconds, scores better than three rushed points in 20 seconds.

Speaking is the section most students feel least prepared for — partly because it's hard to practice alone, and partly because hearing your own voice in a second language is uncomfortable. That discomfort gets smaller the more you record yourself. Start there, and the rest of the prep falls into place.

Share this guide: