TOEFL iBT · January 2026 format
On 21 January 2026 the TOEFL iBT dropped the 0–120 score, the integrated tasks and the independent essay. What replaced them is shorter and denser: 120 items in about ninety minutes, Reading and Listening adaptive, and each section reported on a 1–6 band scale aligned to the CEFR. Preparation written for the old test is preparation for an exam that no longer exists.
In ETS’s order — Reading, Listening, Writing, Speaking — and with the behaviour that actually catches people out, which is rarely the questions.
Two-stage and adaptive: Complete the Words, everyday texts and an academic passage — 20 items a module, and you may move freely inside one.
Single responses, conversations, announcements and academic talks. Each recording plays once, you cannot skip a question, and there is no going back.
Ten Build a Sentence items, an email, and a written contribution to an academic discussion — the last two scored 0–5 on ETS's published criteria.
Seven sentences to repeat and a four-question interview, recorded against the clock with no preparation time.
Reading and Listening are marked the moment you finish, weighted by item difficulty so the easier and harder second modules stay comparable. Writing and Speaking are judged on ETS’s published criteria and returned out of five, each with the reasoning.
Task fulfilment, development, range and accuracy — each scored separately.
Eleven responses against the real clock, transcribed and scored on what you said.
Reading and Listening marked item by item, with why each answer was right.
The 1–6 half-band scale and its CEFR alignment are ETS’s, and so is the rule that the overall score is the mean of the four sections rounded to the nearest half band.
| Band | CEFR | What it describes |
|---|---|---|
| 6 | C2 | Reads and follows anything in the academic register, including argument that turns on nuance. |
| 5.5 | C1 | Handles full-length lectures and dense text; expresses a position with support and few errors. |
| 5 | C1 | Copes with academic study in English, with occasional gaps under time pressure. |
| 4.5 | B2 | Follows the main line of a lecture and writes clearly, but loses detail at speed. |
| 4 | B2 | Manages familiar academic topics; unfamiliar ones cost accuracy or completeness. |
| 3.5 | B1 | Understands the gist of everyday and simple academic material. |
| 3 | B1 | Handles routine exchanges and short factual texts. |
The band descriptions here are ours, written to be useful rather than official, and so is the CEFR column: ETS says the scale is aligned with the CEFR and gives a CEFR range per task type in its blueprint, but has not published a band-by-band table or descriptors. Turning a raw score into a band is also our estimate: ETS equates scores against a calibrated item pool and publishes no conversion table. Your raw score is the fact.
Institutions are still republishing their requirements against the new scale, so check the number your university asks for rather than converting from an old 0–120 target. How the 1–6 scale works →
Yes. The test changed on 21 January 2026: the 0–120 scale, the integrated tasks and the independent essay are gone. This practice test follows the published 2026 specification — 120 items, four sections in the order Reading, Listening, Writing, Speaking, and scores reported on the 1–6 band scale.
Yes, and the routing runs on the server. Everyone sits the same first module; your answers to it decide whether the second module leans easier or harder. You are never sent the module you did not take, and the report says which branch you sat — because the same raw score means different things after the easier and the harder module.
The 1–6 half-band scale, its CEFR alignment and the rule that the overall is the mean of the four sections rounded to the nearest half band are all ETS's. Turning a raw score into a band is our estimate: ETS equates scores against a calibrated item pool and publishes no conversion table. Your raw score is the fact; treat the band as a guide.
About ninety minutes of test time, plus the direction screens between sections. That is close to the real sitting, and it is short enough that pacing errors show up immediately rather than being absorbed by a long paper.
No, and they never will be. Every passage, recording, prompt and question here is written by ExamStep. The format, the timing and the on-screen behaviour are reproduced from the published specification and the official practice interface; the content is our own.
Preparing for a Canadian immigration programme instead? CELPIP practice is live with four free tests.