How An IELTS Band Score Is Built From Four Section Marks

Put a Test Report Form in front of most candidates and they will read one number on it. Usually the wrong one.

The form carries five. Four of them were awarded and one was calculated, and almost every argument about IELTS scoring comes from treating the calculated one as if it were awarded. Read the form from top to bottom and four questions answer themselves: what each number is, how it was produced, what the owners say it means, and how long it lasts.

Everything quoted below is the owners’ own wording, including the three worked examples. If you are not yet sure which version of IELTS you are sitting, settle that first, because the band scale is the one thing all of them share. Where the owners do not publish something, that is said plainly rather than filled in from somewhere else.

Your Test Report Form Carries Five Numbers And Only One Of Them Is Calculated

The owners describe the form in one sentence on their scoring page: “When you receive a Test Report Form you will see your overall test score as well as your scores for the Listening, Reading, Writing, and Speaking sections of the test.”

So four section bands and one overall band. The distinction between them is the single most useful thing to grasp, because the four are results and the fifth is arithmetic performed on those results.

Read the last row again. Nobody decided your overall band. It fell out of a calculation, which means it can be predicted before results day and rebuilt afterwards from the four numbers you already hold.

That also explains why two candidates with the same overall band can be in completely different positions. The overall hides the shape of the four, and plenty of organisations read the shape rather than the summary. Which paper you sat changes the shape too, because the two Reading papers do not convert marks identically.

Every Band From 0 To 9 Carries A Published Skill Level

Scores are “reported in whole and half bands” from 0 to 9, and the owners publish a skill level and a descriptor for each whole band. The labels are worth reading carefully, because the versions circulating on preparation sites are not the ones on the scoring page quoted above.

Three things fall out of that table that are worth more than the table itself.

•   The owners attach the word user to one label only. On the scoring page the five labels candidates quote most read Expert, Very good, Good, Competent and Modest. The circulating forms that append user to each of those are not on that page. Band 1 is the single place the word appears at all, as Non-user, and band 0 is the only label that runs to a phrase rather than a word or two.

•   Band 7 is described as operational, not fluent. Band 8 and band 9 share “fully operational command” while band 7 drops to “operational command” with “occasional inaccuracies”. A candidate chasing a 7 is not chasing perfection, and preparation built on that misreading wastes months.

•   Half bands have no descriptor. The scale reports halves but the published descriptors cover whole bands only. There is no official account of what a 6.5 means, which is why a destination asking for 6.5 is asking for a threshold rather than for a described ability.

That last point has a practical edge. A destination asking for 7.0 when you hold 6.5 is asking you to cross a threshold the owners have never described, so there is no descriptor to study your way across. Why candidates plateau at 6.5 is our account of what actually sits in that half band.

The Overall Band Is An Average Rounded To The Nearest Half Band

The rule is one sentence: “The overall band score is the average of the four section band scores rounded to the nearest half band.”

The owners then publish three worked examples, and these are theirs rather than ours. We have checked the arithmetic in all three and it holds.

Test taker A is the one to study. A 5.0 in Writing sits four half bands below the Speaking score, and the overall still reports 6.5. Nothing on the form flags that gap, and an organisation reading only the overall would never see it.

Test takers B and C move in opposite directions under the same rule. An average of 3.875 reports as 4.0, which is upward, because 4.0 sits 0.125 away and 3.5 sits 0.375 away. An average of 6.125 reports as 6.0, which is downward, for the mirror reason. Neither is generous or harsh, and neither direction is a policy. The nearer half band wins and nothing else is applied.

One practical use for this. Before results day you can estimate all four sections, average them, and know your overall within a half band. Before a resit you can test whether lifting one section would actually move the overall, which it often does not. That single calculation is the cheapest piece of preparation advice available, and our IELTS classes in Chennai start with it rather than with a syllabus.

Both Rounding Edge Cases Move Upward

An average can land exactly between two half bands. There are two such cases, and the owners resolve both in the candidate’s favour in a sentence printed further down the same scoring page, after the worked table.

“If the average of the four sections ends in .25, the overall band score is rounded up to the next half band, and if it ends in .75, the overall band score is rounded up to the next whole band.”

Both move up. That is the detail reported backwards most often, and it has a real cost: a candidate told that .25 rounds down will book a resit they do not need.

Test taker A above is the .25 case in the owners’ own example: an average of 6.25 reporting as 6.5.

•   A warning if you check this in a spreadsheet. Several rounding functions resolve a tie to the nearest even number, which would turn 6.25 into 6.0 and 7.25 into 7.0. That is the opposite of the published rule. Do the two edge cases by hand.

Listening And Reading Convert 40 Marks Into A Band

These two sections are counted rather than judged. The owners state it the same way for both: the test “contains 40 questions”, “each correct answer is awarded 1 mark”, and “scores out of 40 are converted to the IELTS 9-band scale”.

So there must be a conversion, and the owners publish part of it. Not all of it: the figures they give are the “average number of marks scored at different levels” and they cover four bands per paper, not nine. There is nothing published for band 9, nothing for any half band, and nothing below band 5 except on one paper. Here is the whole of it, in one place, with the gaps left as gaps.

Read the band 7 row across. Thirty marks on Listening, thirty on Academic Reading, thirty-five on General Training Reading. That is a five-mark difference on the same band, on the same scale, and the owners state the reason outright rather than leaving it to be inferred: the two Reading tests are “graded on the same scale”, the Academic paper “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”.

So the General Training Reading paper is the one that needs more right answers for the same band, not fewer, which is the opposite of what candidates are usually told when they are advised that General Training is the easier paper. Which paper you are sitting decides which column applies to you, and how the Reading section rewards technique covers closing the gap.

One caveat sits on every number in that table, and it matters more than the numbers: “the precise number of marks needed to achieve these band scores will vary slightly from test version to test version”. These are averages across versions, not a pass mark for the paper in front of you. Treat the figure for your target band as a floor to clear by a margin rather than a target to hit exactly, which is why how our classes are structured has candidates working above their required band rather than at it.

Your Writing Band Is Two Averages Deep And Your Speaking Band Is One

There is no mark out of 40 in these two. The owners’ test format pages set out the tasks; the scoring page sets out how they are judged. Examiners award against criteria, and the averaging happens more than once, which is why a Writing band can feel disconnected from any single thing you did.

For Writing the owners state that examiners “use assessment criteria to award a score for each of the following four criteria”, then: “Each task is assessed independently. The criteria are weighted equally and the score on the task is the average. The assessment of Task 2 carries more weight in marking than Task 1.”

Read that as two averaging steps. Four criteria are scored and averaged to give a score for each task. The two task scores are then combined, with Task 2 weighted more heavily. Only that combined figure reaches your form: the eight criterion scores and the two task scores behind it are never printed anywhere you can see them.

Speaking has one averaging step rather than two, because there are no separate tasks to combine. The owners name its four areas as fluency and coherence, lexical resource, grammatical range and accuracy, and pronunciation, and state that “each of the criteria carry equal weighting and the overall average gives the IELTS score for Speaking”.

The Task 2 weighting is the actionable half of this, and it is why a Task 2 built for band 7 earns more than the same hours spent on Task 1. For Speaking, what examiners are listening for works through the four areas one at a time.

The Owners Recommend Two Years And Say Organisations May Accept Longer

Two years is the figure every candidate has heard. The wording behind it, on the same scoring page, is more conditional than the figure suggests, and the condition is worth money.

“We recommend that IELTS test results are considered valid for two years after a test has been taken. This is based on research into second-language loss. Organisations that accept IELTS can choose to accept results for a longer period, so you should check with them directly.”

Three things follow. It is a recommendation rather than a mechanism, so nothing switches off on the anniversary of your test. It rests on a stated reason, which is research into second-language loss rather than administrative convenience. And the organisation reading your result may accept it for longer, which means a certificate you have written off might still work.

That is not permission to relax. A destination can also apply a window shorter than two years, and some do, so the recommendation is a default rather than a ceiling or a floor. The only reliable answer comes from the organisation you are applying to.

•   Worth asking before you pay. If your result is a little over two years old and your destination has not published a window, ask them in writing before booking a resit. Send us the dates and the destination if you would rather we read their published wording first.

The Owners Publish A Band To CEFR Mapping, With Caveats

Band to CEFR level conversions are everywhere, usually as a confident grid with no source attached. The CEFR is not the only framework a band gets converted into either: Canada converts the same four bands into a Canadian Language Benchmark and pays points for the level rather than the band. The owners publish one too. What the copies drop is everything the owners say around it, and that is the part worth reading.

It sits in two places. Their comparison page carries a CEFR to IELTS band score chart as a figure, and links to a page of their own called IELTS and the CEFR where the mapping is explained. That page is blunter about the limits than any third-party grid.

Start with what the owners say the mapping is not. The framework “was not designed to provide the basis for precise equating, nor was it intended to be a prescriptive tool to impose standardised solutions”. And IELTS does not fit it cleanly: the test “is not a level-based test, but rather designed to span a much broader proficiency continuum”, it uses a nine-band scale, and “thus, there will not be a one-to-one correspondence between IELTS scores and CEFR levels”.

Then the sentence that explains something candidates in India ask about constantly. Why does one university want 6.5 and another 7.0 for what looks like the same course? Because of where the owners’ research puts the C1 boundary.

“our research shows that a C1 minimum threshold would fall between the 6.5 and 7 bands on the IELTS scale. Therefore, whilst many 6.5 test takers would be at C1, a number will be marginally below. So, if an institution requires a high degree of confidence that an applicant is at C1, they may wish to set a requirement of 7, rather than 6.5.”

Read that as the owners advising institutions to ask for 7.0 when they want certainty about C1. The half band between 6.5 and 7.0 is not a scoring quirk. It is the margin the owners themselves identified, and it is why the requirement you are chasing may be a half band above where your ability actually sits. What that half band actually contains is our account of it from the classroom rather than from the research.

Three more things on that page are worth having before you quote any grid.

•   The top of the scale is specific. On whether IELTS differentiates at C2, the owners answer: “Band scores of 8.5 and higher are recognised as C2. Band 8 is borderline.” So a destination asking for C2 is asking for 8.5, not 8.0.

•   The mapping is of overall bands only. The owners state that the band scores in their figure “are the overall band scores, not the individual band scores for Listening, Reading, Writing, and Speaking”. Any grid applied to a single section is being used for something the owners did not build it for.

•   Borderline is their word, not a hedge of ours. They acknowledge it directly, noting that “IELTS band scores have never aligned exactly with the CEFR transition points” because IELTS preceded the framework. Band 5 sitting ambiguously between B1 and B2 is their own worked example of it.

The owners also publish a second mapping, and it works the other way round. Their UKVI page states which band satisfies which CEFR level for UK visa routes, which is a requirement rather than an equivalence. Those figures belong with the test types, and our what is IELTS page carries them.

Note what that is and is not. It is the UK government’s requirement expressed in CEFR terms for one set of visa routes. It is not a general equivalence between the IELTS scale and the framework, and it should not be lifted out of its context and applied to a university or an employer.

The same page does print one conversion candidates rarely know is there, and it is per section rather than overall. Canadian Language Benchmarks, which Canadian immigration applications are expressed in, are mapped to IELTS section bands on the owners’ comparison page. The top of their table runs as follows.

Notice that the four columns are not identical. At CLB 9 the Listening requirement is 8.0 while Reading and Speaking are 7.0, so a candidate averaging comfortably above the line can still miss on one section. This is the per-skill problem in its most expensive form, and the next section is about nothing else.

•   The same page carries four cross-test grids, which is more than most candidates realise: Pearson PTE Academic, Canadian Language Benchmarks, the Cambridge English Scale and TOEFL iBT. The TOEFL one is dated explicitly, applying only to tests taken before January 2026. So the IELTS to PTE equivalence is the owners’ own figure rather than a third-party estimate, and how PTE converts to IELTS bands is where we work through it for candidates deciding between the two.

An Overall Band And A Per-Skill Minimum Are Two Different Tests Of One Result

This section decides whether the rest of the page is any use to you. A requirement normally comes in two parts. Candidates prepare for the first part and are stopped by the second.

Look at test taker A in the owners’ own example again: an overall 6.5 with a 5.0 in Writing. Against a requirement of 6.5 overall that result passes. Against 6.5 overall with no band below 6.0 it fails, on the same certificate, with nothing on the form drawing attention to it.

Where a single section is the problem, a full resit is usually the wrong instrument, because it puts all four bands back at risk to fix one. The owners’ One Skill Retake exists for exactly this, and it is less risky than candidates assume because the retake produces a second certificate rather than overwriting the first. Three limits bind it: the original has to have been an IELTS on computer test at a centre that offers the retake, you get one retake per original test, and it must be sat within sixty days of that test day rather than of your result. How One Skill Retake works and who refuses it covers the conditions and the acceptance problem, which is the part that decides whether it helps you at all.

Which band and which minimums apply is set by the organisation reading your result and by nobody else. Healthcare candidates have a second thing to check, because many regulators also read the profession-specific alternative. For a worked example of a requirement expressed per section rather than as an average, the UK nursing regulator is the clearest one we have written up.

To run this against your own certificate rather than in the abstract, give us the four bands and the destination. We will say which half of the requirement is stopping you, and whether a single-section retake would be enough. Our Chennai and Pondicherry batches are built from that answer, and our healthcare recruitment team takes over once registration is settled.

What Candidates Ask About IELTS Band Scores

What is a good IELTS band score?

There is no universally good score, because IELTS has no pass mark. The owners publish a skill level for each whole band: band 9 is Expert, 8 is Very good, 7 is Good, 6 is Competent and 5 is Modest. What counts as good for you is whatever the organisation reading your result requires, which is usually an overall band plus a minimum in each section. Band 7 is described by the owners as operational command with occasional inaccuracies, not as fluency.

How is the IELTS overall band score calculated?

It is the average of the four section band scores, rounded to the nearest half band. The owners publish three worked examples: 6.5, 6.5, 5.0 and 7.0 averages 6.25 and reports as 6.5; 4.0, 3.5, 4.0 and 4.0 averages 3.875 and reports as 4.0; and 6.5, 6.5, 5.5 and 6.0 averages 6.125 and reports as 6.0. The overall band is the only number on the Test Report Form that nobody awarded, because it falls out of the arithmetic.

Does an IELTS average of 6.25 round up or down?

Up, to 6.5. The owners state that an average ending in .25 is rounded up to the next half band, and an average ending in .75 is rounded up to the next whole band. Both edge cases move upward. This is the detail most often reported backwards, and getting it wrong can send a candidate to a resit they do not need. Be careful checking it in a spreadsheet, because several rounding functions resolve a tie to the nearest even number instead.

What does each IELTS band score mean?

The owners publish a skill level and a descriptor for every whole band from 0 to 9: 9 Expert, 8 Very good, 7 Good, 6 Competent, 5 Modest, 4 Limited, 3 Extremely limited, 2 Intermittent, 1 Non-user and 0 Did not attempt the test. Bands 8 and 9 share the phrase fully operational command, while band 7 is described as operational command with occasional inaccuracies and misunderstandings in some situations. Half bands have no published descriptor.

How many marks do I need for band 7 in Listening or Reading?

The owners publish average figures: 30 out of 40 in Listening, 30 in Academic Reading and 35 in General Training Reading. They attach a caveat that the precise number of marks needed will vary slightly from test version to test version, so these are averages across versions rather than a pass mark for your paper. Note the five-mark gap on General Training Reading, which the owners explain by saying it is usual that a greater number of questions must be answered correctly on that paper to secure a given band score. Aim above the figure for your band rather than at it.

How is the IELTS Writing band score calculated?

Examiners award against four criteria. The owners state that each task is assessed independently, that the criteria are weighted equally and the score on the task is the average, and that the assessment of Task 2 carries more weight in marking than Task 1. So four criteria are averaged to produce a score for each task, and the two task scores are then combined with Task 2 weighted more heavily. Only that final figure appears on your Test Report Form.

How is the IELTS Speaking band score calculated?

Examiners award against four areas: fluency and coherence, lexical resource, grammatical range and accuracy, and pronunciation. The owners state that the four criteria carry equal weighting and that the overall average gives the Speaking score. Because the weighting is equal, no single area can be neglected, which is different from Writing where a weak Task 1 is partly offset by a stronger Task 2.

How long is an IELTS band score valid?

The owners recommend that results be considered valid for two years after the test, and they say this is based on research into second-language loss. It is a recommendation rather than an expiry: they add that organisations accepting IELTS can choose to accept results for a longer period and that you should check with them directly. A destination can also apply a shorter window, so the only reliable answer comes from the organisation reading your result.

What is the IELTS equivalent of CEFR B2 or C1?

The owners publish a mapping, on their comparison page as a chart and on their IELTS and the CEFR page in prose, but they attach caveats most copies drop. They state there will not be a one-to-one correspondence between IELTS scores and CEFR levels, because IELTS is not a level-based test. On C1 specifically their research puts the threshold between the 6.5 and 7 bands, so they advise an institution wanting high confidence in C1 to set 7 rather than 6.5. Band scores of 8.5 and higher are recognised as C2, with band 8 described as borderline. And the mapped figures are overall band scores, not individual section bands.

Can I pass IELTS overall and still be rejected?

Yes, and it is common. Most requirements have two parts, an overall band and a minimum in each section, and averaging cannot help with the second. The owners’ own worked example shows a candidate with an overall 6.5 whose Writing section is 5.0. That result satisfies a requirement of 6.5 overall and fails a requirement of 6.5 overall with no band below 6.0, on the same certificate. Read both halves of any requirement before you book.

Should I resit the whole test to fix one section?

Usually not. A full resit puts all four bands back at risk in order to repair one. The owners offer IELTS One Skill Retake for this situation, which replaces a single component and produces a new Test Report Form. Before you choose it, two things are already fixed by your original booking: the sitting had to be on computer, at a centre that offers the retake. Separately, your destination has to accept the resulting form. Then the clock. Sixty days from your original test day, once only, so the decision is made in the weeks after your result rather than whenever you feel ready. Weigh that against a full resit, which risks four bands to repair one, and note that the retake produces a second certificate while leaving your original intact. Our full account of the conditions sits on our One Skill Retake page.

Do half bands like 6.5 have a meaning?

No descriptor is published for them. The scale reports results in whole and half bands while the owners’ descriptors cover whole bands only, so there is no official account of what a 6.5 describes. The one place the owners do give a half band meaning is on CEFR: their research puts the C1 threshold between 6.5 and 7, and they advise institutions wanting confidence in C1 to ask for 7. So the gap between 6.5 and 7.0 is a real margin in the owners’ own research rather than a rounding artefact.