TOEFL Speaking: Task by Task
TOEFL Speaking is short: about 16 minutes, four tasks, one microphone. You hear prompts through headphones and speak into a microphone. ETS scores each response from 0 to 4, then scales the total to 0-30. Scoring uses human raters plus AI.
For official sample tasks and scoring guides, visit the ETS Speaking section.
"You Understand English But Can't Speak" — clear pace, practical
"Daily English Speaking Practice" — built for repeating
"No Partner? No Problem!" — three ways to practice alone
The four tasks
- Task 1 — Independent opinion: 15 seconds to prepare, 45 seconds to speak. You answer a personal choice question such as "Do you prefer studying alone or with friends?"
- Task 2 — Campus situation: Read a short notice, listen to a conversation, then summarize one speaker's opinion. 30 seconds prep, 60 seconds speaking.
- Task 3 — Academic concept: Read a short academic text, listen to a lecture, then explain the concept using lecture examples. 30 seconds prep, 60 seconds speaking.
- Task 4 — Academic lecture: Listen to a lecture only, then summarize it. 20 seconds prep, 60 seconds speaking.
Tasks 2 and 3 are read-listen-speak. Task 4 is listen-speak. Task 1 is your own opinion only.
Scoring in plain words
Each response gets 0 to 4. Three dimensions decide the score.
- Delivery: how clearly you speak and how natural the pacing is. Long silent pauses hurt. You do not need a native accent, but you must be easy to follow.
- Language Use: grammar and word range. Repeated simple sentences and basic verbs hold the score down.
- Topic Development: whether you answer the exact question and support your reasons. This is where scores often stop growing.
Templates help, then they hurt
A template is a skeleton. Your content is the flesh. A memorized skeleton gives you a start, but a recited template caps your score because every answer sounds the same.
Here is a plain Task 1 skeleton:
I think ___ because ___. First, ___. For example, ___. Second, ___. That is why I choose ___.
That is enough. The problem starts when only the nouns change. The rater hears the frame and not the person.
The fix is practice. Decide your two reasons and one example before you speak. Then speak them through the skeleton until the skeleton is invisible. By test day, the words should bend around your content, not the other way.
Notes for integrated tasks: two columns
For Tasks 2 and 3, draw a vertical line down your note page. Left side: the reading point. Right side: the lecture point.
| Reading point | Lecture point |
|---|---|
| University will end evening bus service | Student disagrees; many art students rely on it |
| Reason: low ridership | Ridership is low because the route is unreliable, not useless |
Task 3 uses the same layout: left side is the academic term and its definition, right side is the lecture example that makes it concrete. Task 4 has no reading, so use one column with the main concept on top and examples below.
The 45-second discipline
Forty-five seconds is short. Sixty seconds feels longer, but it also disappears. Plan two points, not three. A finished answer with two reasons beats an unfinished answer with three.
If your timer says 38 seconds and you are about to start a third reason, stop. Finish the sentence you are in and add a short closing. An unfinished last sentence costs more than a missing third reason. It tells the rater you lost control of time.
Practice with a phone timer until 45 seconds feels like enough, not a trap.
Microphone anxiety
Speaking into a microphone is different from speaking to a person. The silence around your voice feels heavy. The fix is to make it normal. Practice with your phone's voice recorder, not just in your head. Record one 45-second answer, then listen back and count your pauses. Your voice will sound odd to you. That is not a stop sign.
Two guides help here: record yourself speaking English and overcoming English speaking anxiety. Do not let test day be the first time you hear yourself through a mic.
A three-week plan
| Week | Focus | Daily work | Time |
|---|---|---|---|
| 1 | Format and Task 1 | Read the four task types. Answer two Task 1 questions aloud into a recorder. Listen back once. | 20 minutes |
| 2 | Integrated tasks | Do one Task 2 and one Task 3 per day. Take two-column notes. Time every answer. | 30 minutes |
| 3 | Full sets and delivery | Do a full four-task set every other day. On off days, shadow for 10 minutes and review your recordings. | 35 minutes |
Where shadowing fits
Shadowing means speaking along with a video a fraction behind the speaker. For theory and variants, see what is English shadowing. For TOEFL-specific delivery drills, use shadowing for IELTS and TOEFL.
Ten minutes a day of shadowing polishes delivery: pacing, sound, and mouth movement. It will not build your Task 3 content. Do it after one real speaking task, not instead of one.
Score plateaus
Score plateaus after 22 or 23 usually come from content thinness, not accent. Many learners blame their accent. Usually that is not the wall. A clear foreign accent can still reach 26 if the answer is full and complete. A native-like accent with empty reasons will not.
If your score stops, record two answers and listen only for content. Did you answer the exact question? Did you give two reasons with examples? Did you finish every sentence? That is where the points are.