Score 30 out of 40 in IELTS Reading and your band depends on which test you sat. On the official ielts.org tables, 30 marks is Band 7 on Academic Reading — and Band 6 on General Training Reading. Same raw score, one whole band apart. That is not a prep-site theory — it is printed on the IELTS partners' own scoring page. Their own explanation is hedged, not blunt: Academic texts "may contain texts which feature more difficult vocabulary or greater complexity of style" — so GT asks for more correct answers to earn the same band.
If you are practising with Academic materials for a GT test, or judging your GT mock by an Academic chart, you are grading yourself a band too generously. This page prints the tables officialdom actually publishes — all of them — and tells you plainly where no official table exists at all.
The failure
The mistake this page prevents: treating a full 0–40 raw-to-band conversion chart as official, for any skill. There isn't one. ielts.org publishes only partial "average marks" tables for Listening and the two Readings — four anchor rows per skill, nothing more — with an explicit warning attached:
"Here are the average number of marks scored at different levels of the IELTS scale in Listening section. The precise number of marks needed to achieve these band scores will vary slightly from test version to test version."
The Reading tables carry the same "will vary slightly from test version to test version" framing. So even the official numbers are averages, not cut-offs. A complete chart that fills in every band from 1 to 9 — the kind most prep sites display — is a fabrication built around these four rows.
The official tables — everything that exists
Listening and Reading each have 40 questions, one mark per correct answer. IDP states the mechanism:
"The Listening and Reading tests contain 40 questions and each correct question will be awarded 1 mark (so the maximum a test taker can score here is 40). Band scores, ranging from band 1 to band 9, are awarded based on the raw scores."
Here are the average marks ielts.org publishes, in full. This is the entire official dataset:
| Band score | Listening | Academic Reading |
|---|---|---|
| 8 | 35 | 35 |
| 7 | 30 | 30 |
| 6 | 23 | 23 |
| 5 | 16 | 15 |
| GT Reading band | GT Reading marks |
|---|---|
| 7 | 35 |
| 6 | 30 |
| 5 | 23 |
| 4 | 15 |
Read the GT table against the others and the headline fact appears: the marks that average Band 8, 7, 6 and 5 on Academic Reading average Band 7, 6, 5 and 4 on General Training Reading. The conversion is systematically one band lower on GT for the same raw score. The official reason stops short of "easier": Academic texts "may contain texts which feature more difficult vocabulary or greater complexity of style", and "it is usual that a greater number of questions must be answered correctly on a General Training Reading test to secure a given band score."
There is a second, smaller asymmetry hiding in one cell: Band 5 averages 16 marks in Listening but 15 in Academic Reading. The tables are near-identical, not identical — another reason to quote the exact cells rather than a remembered "same scale".
A worked example: the GT candidate who thought 30 was enough
You are preparing for General Training and, in a timed mock, you score 30/40 in Reading.
- Against the Academic column, 30 marks reads as Band 7. Most conversion charts online show the Academic figures without labelling them, so this is the number you will probably find first.
- Against the official GT table, 30 marks averages Band 6. One whole band lower — the score that actually corresponds to your test.
- Even the GT figure is an average, not a guarantee. The official caveat means the precise cut-off on your test date can sit slightly either side of 30.
- To average Band 7 on GT Reading, the official table says you need around 35 marks — the same raw score that averages Band 8 on Academic. That is five more correct answers than you just got, not zero.
The practical rule: mark GT Reading mocks against the GT column only, and treat the result as an estimate with a margin, not a cut-off. Listening needs no such adjustment — the Listening test is the same on Academic and GT, and it has one table.
Writing and Speaking: no table exists, official or otherwise
There is no official raw-score-to-band conversion for Writing or Speaking. None. The ielts.org scoring page, the British Council's results-explanation page, IDP's scores page and Cambridge's test-format page were all checked on 24 August 2026 — no such table appears on any of them.
That is because there are no raw marks to convert. Writing and Speaking are examiner-marked against four criteria each — for Writing: task achievement (Task 1) / task response (Task 2), coherence and cohesion, lexical resource, and grammatical range and accuracy; for Speaking: fluency and coherence, lexical resource, grammatical range and accuracy, and pronunciation. An examiner rates your performance against band descriptors — there is no answer sheet, no /40, and so nothing for a table to map.
Every confident "Writing raw score chart" or "Speaking percentage table" on the web is invented. If a site shows you one, it has told you something about its accuracy on everything else too. On this topic the honest answer is short: for Listening and Reading, four official anchor rows with a variance caveat; for Writing and Speaking, descriptors and an examiner.
Once all four section bands exist, the overall band is their average rounded to the nearest half band — an average ending in .25 rounds up to the next half band, .75 up to the next whole band. That calculation has its own page in the related links.
What actually costs you
- Marking a GT Reading mock with an Academic chart. You will walk into the test believing you are a band better than you are. 30/40 is not Band 7 on GT.
- Treating any anchor row as a cut-off. The official wording is "average number of marks" that "will vary slightly from test version to test version". A 29 is not certainly a miss, and a 30 is not certainly safe.
- Trusting a filled-in 0–40 chart. Only the rows printed above are official. Bands 1–4 and 9, and everything between the anchors, are somebody's guess.
- Planning a Writing band from a "raw score". Writing has no raw score. Time spent decoding fake tables is time not spent on the four criteria examiners actually use.
- Forgetting which Reading test you booked. Speaking and Listening are the same on Academic and GT — Reading and Writing are the different ones, and Reading's scoring differs too.
Try one
The only conversion that matters is the one applied to answers you actually produced under exam timing — sit a full test and see where your raw scores land against the official rows.
[Try a full mock test →] — one full task, marked, no card required.