What does the TOEFL iBT Speaking section test?
The TOEFL iBT Speaking section tests your ability to speak English in academic contexts through 4 tasks completed in approximately 16 minutes. One Independent task asks you to express and defend your opinion, while three Integrated tasks require you to synthesize information from reading and listening passages before speaking.
The TOEFL iBT Speaking section is where many high-performing test-takers lose points they cannot afford to lose. A candidate might score 28 on Reading, 27 on Listening, and 26 on Writing — then receive a 21 on Speaking because they did not understand what ETS raters are actually evaluating. This guide explains the section with precision, covering task structures, scoring criteria, what raters hear in a 24-26 response versus an 18-20 response, and how to build speaking skills that hold up under test conditions.
Task Structure and Timing
The Speaking section runs approximately 17 minutes and consists of exactly 4 tasks. Understanding the timing of each task is not optional — running over or under your allotted time is one of the most common reasons responses score poorly.
| Task | Type | Preparation Time | Response Time |
|---|---|---|---|
| Task 1 | Independent | 15 seconds | 45 seconds |
| Task 2 | Integrated (Campus) | 30 seconds | 60 seconds |
| Task 3 | Integrated (Academic) | 30 seconds | 60 seconds |
| Task 4 | Integrated (Academic Lecture) | 20 seconds | 60 seconds |
Task 1: Independent Speaking
You are given a question asking for your opinion, preference, or recommendation on a familiar topic. Topics range from whether students should live on campus or off campus, to whether working while studying is beneficial, to what quality is most important in a leader. You have 15 seconds to prepare and 45 seconds to respond.
The key word in the task instructions is "defend." ETS is not looking for you to state a preference and then stop. You need to state a position clearly, provide at least one reason, and support that reason with a specific detail or example. Forty-five seconds is enough time to do this exactly once with sufficient development, or to do it twice briefly. Most high-scoring responses state one point and develop it fully rather than rushing through two underdeveloped points.
Tasks 2-4: Integrated Speaking
Tasks 2, 3, and 4 all require you to read a short passage, listen to a spoken segment, and then produce a response that synthesizes the two sources. The reading passages in Tasks 2 and 3 are visible during the listening portion but disappear before you speak. Task 4 has no reading passage — only a lecture.
Task 2 involves a campus announcement or policy change. The listening portion features a student conversation about the announcement. Your job is to explain the announcement and how the speakers react to it, including their reasons.
Task 3 involves an academic term or concept defined in a reading passage. The lecture then provides one or two specific examples of that concept. Your job is to explain the concept and how the lecture examples illustrate it.
Task 4 involves an academic lecture on a topic from a particular field. The lecture explains a concept or process using examples. You summarize the lecture's main points and the examples used to support them.
What ETS Raters Score: The Three Criteria
Every TOEFL Speaking response is scored by a trained human rater using three criteria. Understanding these criteria precisely changes how you practice and how you structure your responses.
1. Delivery
Delivery covers the clarity and naturalness of your speech. Raters evaluate your pace (are you speaking too fast or too slow?), your pronunciation (can your words be understood?), your intonation (do you sound monotone or robotic?), and your fluency (are there long pauses, excessive filler sounds, or false starts?).
A common misconception is that Delivery rewards a "native-sounding" accent. ETS explicitly states that accent does not affect scoring. What matters is intelligibility — can a trained English speaker understand you without difficulty? A French-accented speaker who speaks clearly and at an appropriate pace will score as high on Delivery as a native speaker.
2. Language Use
Language Use evaluates your grammatical accuracy and the range of vocabulary you use. Raters are not looking for perfect grammar — they are looking for the ability to convey meaning accurately despite the grammatical errors that appear in natural speech. However, patterns of grammatical error that impede meaning will lower your score.
Vocabulary range matters more than vocabulary size. Using the same three sentence structures repeatedly signals a limited repertoire even if each sentence is grammatically correct. High-scoring responses use subordinating clauses, relative clauses, conditional structures, and a range of discourse markers naturally.
3. Topic Development
Topic Development is the most misunderstood criterion. It evaluates whether your response is coherent, complete, and appropriately connected to the question or prompt. Raters ask: Did you address the task? Is there a clear main idea? Are supporting points connected to the main idea? Does the response feel finished, or does it end abruptly?
For Integrated tasks specifically, Topic Development includes whether you accurately captured the content of the reading and listening sources and whether you clearly expressed the relationship between them.
"Raters are trained to evaluate what the test-taker communicates, not how it sounds. A heavily accented response with clear structure and relevant content will consistently outscore a fluent-sounding response that wanders off topic." — ETS TOEFL Scoring Guide, Rubric Notes for Speaking
Scoring: What a 24-26 Response Sounds Like vs. an 18-20 Response
The Speaking section is scored from 0 to 30 in four bands, converted from a 0-4 raw score on each task.
| Score Range | Label | General Description |
|---|---|---|
| 26-30 | Advanced | Clear, mostly fluid delivery; effective language use with few errors; complete and fully developed responses |
| 18-25 | High Intermediate | Some fluency difficulties; grammar and vocabulary mostly effective; ideas mostly clear with some gaps |
| 10-17 | Low Intermediate | Frequent pronunciation or fluency problems; limited language; incomplete or poorly developed ideas |
| 0-9 | Below Basic | Difficult to understand; minimal content; may not attempt task |
A 24-26 Response (Task 1 Example)
The question: "Do you prefer to study alone or with other people? Use details and reasons to explain your answer."
A 24-26 response states a clear preference in the first sentence ("I prefer to study alone"), immediately provides a primary reason ("because I can control my pace and environment"), develops that reason with a specific example ("For instance, when I'm working through difficult statistics problems, I need to spend twenty or thirty minutes on a single concept, and studying with others means we have to move on when most people are ready"), and closes with a brief restatement or implication. The response fills most of the 45 seconds without rushing or trailing off. There are minor grammatical errors but none that interrupt comprehension.
An 18-20 Response (Same Task)
An 18-20 response often starts with the same structure but breaks down in the support. The test-taker states a preference, gives a reason, then either runs out of content and falls silent with 10 seconds remaining, or pivots to a second, underdeveloped point that sounds disconnected. Alternatively, the response is fluent but vague — "I like studying alone because it's more comfortable for me" — with no specific development. Raters can hear the difference between "I have nothing more to say" silence and natural speech pauses.
The Over-Templated Response Problem
Templates are commonly recommended in test prep materials, and they do serve a purpose: they give test-takers a reliable opening and transition structure under pressure. The problem is when templates become the response itself — when a test-taker memorizes 3-4 sentences of template and then cannot fill the remaining time with genuine content.
ETS raters are trained to identify over-templated responses. A response that opens with "This is a very interesting question. There are many different ways to approach this topic" and then delivers thin content will score poorly on Topic Development regardless of how fluent it sounds.
The templates that help are minimal: a one-sentence opener that directly states your position, a one-sentence transition that introduces your reason ("The main reason for this is..."), and a closing phrase that signals you are done ("So, for these reasons, I believe..."). The 35-40 seconds in between must be genuine, specific content.
For Integrated tasks, the useful template is a structural one rather than a verbal one. A reliable structure for Task 3: state the concept from the reading in one sentence, then explain each example from the lecture and connect it explicitly to the concept. This structure takes about 55 of your 60 seconds if you develop each example.
"The most common mistake in TOEFL Speaking preparation is practicing fluency without practicing content generation. Fluency without substance does not score above 22." — Dr. Linda Fang, Applied Linguistics, Georgetown University
How to Practice at Home
Effective Speaking practice requires recording and reviewing your own responses. There is no other way to identify your specific patterns of error.
Step 1: Build a timed practice routine. Use official ETS practice materials (available in the free TOEFL Practice Online samples and in the Official Guide to the TOEFL Test). Set a timer for exact preparation and response times. Do not give yourself extra seconds — the actual test cuts your microphone off.
Step 2: Record every practice response. Use your phone's voice recorder, a laptop microphone, or any available recording tool. The recording quality does not need to be high.
Step 3: Review with the scoring criteria, not your gut. After recording, listen back and score yourself on Delivery, Language Use, and Topic Development separately. Ask: Did I fill the time with relevant content? Did I connect supporting details to my main point? Were there pauses longer than 2-3 seconds? Did I finish the response or stop abruptly?
Step 4: Transcribe 2-3 responses per week. Writing out your spoken response forces you to see grammatical patterns you cannot hear in real time. Repeated grammatical errors become visible when transcribed.
Step 5: Shadow authentic academic English. Use TED Talks, university lecture recordings, or NPR discussions to practice matching the pacing and intonation of educated English speakers. Ten minutes of daily shadowing over 8 weeks produces measurable improvements in Delivery scores.
"Speaking improvement requires feedback. Studying speaking in isolation without ever hearing your own voice is like training for a marathon without ever running." — Margaret Taber, former ETS Speaking rater and TOEFL preparation trainer
Integrated Task Strategies by Task
Task 2 (Campus Situation)
Read the announcement or notice carefully for two things: what the change is, and what the stated reasons for the change are. The conversation that follows will either support or oppose the announcement. In your response, you need to describe the change, state the person's opinion, and explain their reasons. Raters do not want your opinion — they want an accurate summary of the conversation.
Task 3 (Academic Concept)
Read the definition in the passage carefully, particularly any key terms. The lecture will not repeat the definition word-for-word but will provide examples. Your response must explain the concept (in your own words if possible) and explain how each example from the lecture illustrates it. The connector phrase matters: "The professor uses the example of X to illustrate Y" is clearer than just "The professor talks about X."
Task 4 (Academic Lecture)
Task 4 has no reading, which means your entire response is based on listening. The lecture typically has a clear structure: a main idea (often stated explicitly at the start), followed by two examples or two sub-points. Take notes. Your notes do not need to be complete — focus on the main idea and one or two specific examples with their supporting details.
Common Preparation Mistakes
The most significant error test-takers make is focusing exclusively on accent reduction before working on content generation. Pronunciation matters for Delivery, but a speaker with a noticeable accent who delivers coherent, fully developed responses consistently outscores a clear-sounding speaker whose responses are vague or underdeveloped.
The second most common error is practicing only Task 1 while neglecting the Integrated tasks. Task 1 is worth the same as each Integrated task in the final conversion, and many test-takers find the Integrated tasks harder because they require note-taking and synthesis under time pressure.
The third error is ignoring time management. Practicing responses without enforcing the actual time limits creates habits that break down during the real exam. Always use exact timing.
Building a Note-Taking System for Integrated Tasks
For Tasks 2, 3, and 4, note-taking during the listening portion is critical. Notes cannot be reviewed after you start speaking — they are a bridge between the audio and your response. Effective notes for TOEFL Speaking are not complete sentences; they are abbreviated keywords and phrases that trigger memory of the full content when you glance at them during your 30-second preparation.
A practical note structure for Task 3 (Academic Concept):
- Reading note: [Concept name] = [brief definition in 5-7 words]
- Lecture note 1: Example 1 = [key name or term] → [how it illustrates concept in 5-7 words]
- Lecture note 2: Example 2 = [key name or term] → [how it illustrates concept in 5-7 words]
Using these notes during 30-second preparation, you can plan a structured 60-second response: 10 seconds on the concept, 20-25 seconds on example 1 with connection to concept, 20-25 seconds on example 2 with connection to concept, 5 seconds of brief conclusion.
For Task 4 (Lecture only):
- Main point: [topic] and its significance in 5-7 words
- Example/support 1: [key term] + [supporting detail]
- Example/support 2: [key term] + [supporting detail]
The key is developing a consistent note format that becomes automatic. Test-takers who try to take comprehensive notes often write too slowly and miss content; those who take no notes find 60 seconds of organized speaking impossible.
Pacing: The 45-Second and 60-Second Structures
Understanding the architecture of a well-paced response for each task time prevents the two most common pacing errors: ending too early (with 10-15 seconds of silence) or rushing and cutting off mid-sentence.
45-Second Structure (Task 1):
- Sentence 1 (0-5 seconds): State your position clearly
- Sentence 2-3 (5-15 seconds): State your primary reason
- Sentences 4-6 (15-38 seconds): Develop the reason with specific detail or example
- Sentence 7 (38-45 seconds): Brief concluding statement or implication
If you complete the concluding statement before 45 seconds, do not add a second example — rephrase your conclusion or extend your support. Abrupt silence before the cutoff is more damaging to your score than filling time with a natural extension.
60-Second Structure (Tasks 2-4):
Tasks 2 and 3 have more content to cover: the announcement/concept (from reading) and the conversation/examples (from listening). The 60 seconds divides approximately as: 10 seconds overview of situation, 25 seconds for first point/example from listening with connection, 20 seconds for second point/example with connection, 5 seconds conclusion.
Task 4 has no reading and typically involves one main concept from the lecture with two examples. The structure: 10 seconds for the topic, 20-22 seconds for example 1 with detail, 20-22 seconds for example 2 with detail, 6-8 seconds brief conclusion.
Score Conversion and Section Minimums
For competitive university programs, speaking minimums often differ from overall score minimums. Teaching assistant positions in US universities frequently require a minimum TOEFL Speaking score of 23-26 specifically, regardless of overall TOEFL score. Programs with heavy oral presentation requirements set section minimums for the same reason. Check the specific requirements of your target programs — an overall score of 100 with a Speaking score of 20 may not satisfy requirements even if the overall threshold is met.
| Score | Percentile Rank (Approximate) |
|---|---|
| 29-30 | 98-99 |
| 26-28 | 85-95 |
| 23-25 | 65-80 |
| 20-22 | 40-60 |
| 17-19 | 20-35 |
| Below 17 | Below 20 |
References
- Educational Testing Service. (2024). TOEFL iBT Test Content. ETS TOEFL. https://www.ets.org/toefl/test-takers/ibt/about/content.html
- Educational Testing Service. (2023). TOEFL iBT Scoring Guide: Speaking. ETS. https://www.ets.org/content/dam/ets-org/pdfs/toefl/speaking-rubrics.pdf
- Educational Testing Service. (2024). The Official Guide to the TOEFL iBT Test (6th ed.). McGraw-Hill Education.
- Fang, L. (2022). Rubric-based self-assessment in second language speaking. Language Testing, 39(2), 245-268. https://doi.org/10.1177/02655322211034890
- Ginther, A., & Elder, C. (2014). A content analysis of research published in the journals Language Testing and Language Assessment Quarterly. Language Testing, 31(1), 35-61.
- ETS TOEFL. (2024). TOEFL iBT Score Descriptors. https://www.ets.org/toefl/score-users/scores/understand.html
- Xi, X., & Mollaun, P. (2011). How do raters from India perform in the TOEFL iBT speaking scoring process and why? Language Testing, 28(4), 547-576. https://doi.org/10.1177/0265532210394347
- Harding, L. (2011). Accent and listening assessment. Language Assessment Quarterly, 8(1), 72-89.