TOEFL iBT® test · New format
Is your TOEFL prep out of date?
The TOEFL iBT® test was rebuilt in January 2026. New task types, new section order, new timings, and a new 1–6 band score. A great deal of the preparation material in circulation — books, courses, apps, YouTube playlists, question banks — still teaches the previous test. Here is how to check yours in about a minute.
The 60-second test
Forget the marketing on the cover. The current test has twelve task types — but only eight of them are new in 2026, and those eight are what make a useful test. The other four were already familiar in some form from the previous test, so finding them proves nothing:
| Section | New in 2026 — your actual diagnostic | Familiar in some form from the old test |
|---|---|---|
| Reading | Complete the Words · Read in Daily Life | Read an Academic Passage |
| Listening | Listen and Choose a Response · Listen to an Announcement | Listen to a Conversation · Listen to an Academic Talk |
| Writing | Build a Sentence · Write an Email | Write for an Academic Discussion |
| Speaking | Listen and Repeat · Take an Interview | — |
Eight new plus four carried over is the full set of twelve. Note that the carried-over tasks were not left untouched — Listen to an Academic Talk, for instance, uses much shorter talks than the old lectures — so their presence alone does not tell you the material is current.
Complete the Words is the quickest single check. It is a C-test — the first sentence of a passage is left intact, then the second half of every second word is deleted and you type it back — and it accounts for 30 of the 50 items in ETS's published Reading base form. Pre-2026 TOEFL iBT material will not teach it. If your material has never mentioned it, it describes the old format, even if its general English exercises remain useful.
Five more warning signs
- Scores are given out of 120. The current test reports every section and the overall result as a band from 1 to 6 in half-band steps. During the transition, reports also show one comparable overall 0–120 estimate so older university requirements remain usable; they do not add comparable 0–30 scores for each current section.
- The test is described as three hours or more. The 2026 test is about 90 minutes of section work and has no scheduled break. Anything describing a three-to-four-hour sitting with a rest period is several revisions behind.
- There is an "integrated writing" task. The classic read-a-passage, listen-to-a-lecture, write-a-summary task is not part of the current Writing section. If your book drills it, that book is at least one full revision out.
- Speaking is described as four or six tasks. Current Speaking has two task types across 11 items in about eight minutes.
- Adaptivity is never mentioned. Reading and Listening are now two-stage adaptive — your first-stage performance determines the difficulty of the second stage. Material that describes a single fixed set of questions for everyone is describing the old test.
Also check the section order
The current order is fixed: Reading → Listening → Writing → Speaking, with no going back to a completed section. If your practice tests put Speaking third, they are not simulating the test you will sit. This matters more than it sounds: Speaking last, at the end of an unbroken 90 minutes, is a genuinely different experience from Speaking in the middle.
How to check a book, a course or an app
- Books — check the copyright page, not the cover. A cover reading "2026 Edition" is a marketing claim; the copyright year and the presence of Complete the Words are facts. Note that a 2025 copyright date does not automatically mean out of date: ETS announced the changes in May 2025 and began releasing supporting materials from that July, so a book published in late 2025 may well describe the current format correctly. The task names are the reliable test; the date only tells you whether it was possible for the book to be current.
- Apps and question banks — look at the update notes or changelog. Then spot-check: open the Reading practice and see whether a C-test appears at all.
- Video courses — check the upload date on the individual lessons, not the playlist. Playlists get renamed; lessons rarely get re-recorded.
- Free PDFs and shared question banks — these are the worst offenders, because they circulate for years with no version information whatsoever. Apply the task-name test.
- Anything at all — search the document for "Complete the Words", "Build a Sentence" and "Listen and Repeat". Three searches and about a minute. If none of them appear anywhere, that is a strong signal the material predates the current test; if they do appear, check that the timings and score scale match too, since a hasty revision can add the names without fixing the rest.
What is still worth keeping
Do not throw everything away. The distinction that matters is between skills and format.
Still useful: general reading practice, listening to academic English, vocabulary building, grammar work, and any writing or speaking practice aimed at fluency and accuracy. The English being tested did not change.
No longer reliable for the current test: task instructions, section timings, structure, answer strategies tied to retired task types, pacing plans, and practice-score predictions built for the old test. Historical 0–120 university requirements still matter because current reports include a comparable overall estimate, but an old mock cannot predict the new format faithfully.
Why this matters more this year than usual
Test formats normally change gradually enough that slightly old material is only slightly wrong. The January-2026 rebuild was not that kind of change: the item counts, the task types, the section order, the timings, the adaptivity and the score scale all moved together. So material written for the previous test tends to be wrong on most format questions at once rather than a little out of date on one or two — which is why the date on the cover matters less than whether the task types match.
There is also a practical trap. Because current reports include a comparable overall 0–120 estimate during the transition, out-of-date material can look current for longer than it should. Seeing a familiar "aim for 100" target does not mean the book knows what the test looks like now.
If you want the specifics, our section-by-section summary of what changed in 2026 lists every difference, and the format overview covers the current test end to end.
If your material is out of date
- Confirm the format first, before doing any more practice. Practising the wrong task type is worse than not practising, because it builds habits you then have to unlearn.
- Sit one full-length mock in the current format. This tells you two things at once: what the test now looks like, and where you actually stand on it.
- Re-target your score. If your goal was a number out of 120, convert it — our free score converter turns an old target into the band you now need.
- Keep the skills work, replace the format work. Your vocabulary lists are fine. Your pacing plan is not.
Check yourself against the real 2026 format
Take a full-length practice mock built to the January-2026 specifications — all twelve task types, the real order, the real timings — and get a predicted 1–6 band for each section. No card required.
Avermond Test Prep is an independent preparation service. TOEFL® is a registered trademark of ETS. This product is not endorsed or approved by ETS. Format details reflect ETS's published January-2026 TOEFL iBT test materials and may be subject to minor revision; always confirm the current format with the official test provider. This page does not assess or refer to any specific third-party publisher's materials.