How to Know Your IELTS Band Before Booking the Exam

The short answer

You know your likely band before booking by taking a full-length, exam-style mock under exam conditions — all four modules, including writing and speaking, the two that self-study cannot self-score — and getting back an estimated band range aligned to the CEFR scale, not a guessed point score. If the low end of that range meets or beats the band your visa route or university asks for, you book. If it does not, you now know exactly which module is holding you back, and you drill that module before spending exam money.

This matters most where the exam fee is heaviest relative to income. In Nigeria the official test costs upwards of ₦250,000 at 2026 prices; across Nigeria, Kenya, Ghana and South Africa the fee works out to roughly US$180–260 per sitting. Booking on a hunch, missing your band by half a point, and rebooking is the most expensive study mistake in the whole migration journey — and it is entirely avoidable with a US$2 measurement.

Why guessing your band is so expensive

The exam is pass-fail against your target: a 6.5 requirement makes a 6.0 result worth nothing, and the fee is gone either way. Most candidates who miss do not miss by a mile — they miss by half a band in one module, usually writing or speaking, the two they never had scored during preparation. Multiply one avoidable retake by the flight or bus to a test centre, a day off work, and weeks of lost application time, and the real cost of not knowing your band is a multiple of the fee itself.

The candidates who book once and pass once are almost never the ones who studied the longest. They are the ones who measured, found the weak module, fixed it, measured again, and only then paid. Readiness is a number you can know in advance — treating it as a feeling is what the retake industry quietly depends on.

The self-scoring trap: free tests only cover half the exam

Free practice PDFs and answer-key websites score listening and reading well enough — those modules have right answers you can check yourself. But the band you fail on is decided in writing and speaking, and neither can be self-scored: you cannot mark your own essay for task response, coherence, lexical range and grammatical accuracy, and you cannot examine your own two-minute long turn. So most self-study candidates walk into the exam with a solid estimate of half their band and no estimate at all of the half that fails people.

AI scoring closes exactly this gap. A serious mock has you type real essays — a chart-describe task and a discursive essay on the academic paper, a letter and essay on the general-training paper — and record real speech, including a timed long turn, then scores them against CEFR-aligned criteria the way the written and spoken halves of the real exam are assessed. The point is not that AI replaces an examiner on exam day; it is that an AI-scored rehearsal is the only affordable way to have your writing and speaking assessed at all before you pay for the real one.

What an honest band estimate looks like

Be suspicious of any practice test that returns a single confident number — 'your band is 7.0' — from one sitting. No mock, however good, can promise your result on a different paper on a different day, and a tool that pretends otherwise is telling you what you want to hear. An honest estimate is a range: on AssessAll's exam-style mocks the report gives a CEFR level and an estimated band range half a band wide (say 6.0–6.5), because that is the honest resolution of a single measurement, with module-by-module scorecards underneath showing where the range would move.

The same honesty applies to branding: AssessAll's papers are IELTS-style and TOEFL-style — built to the same four-module structure, timing discipline and CEFR alignment — and the store says plainly that it is not affiliated with, endorsed by, or a substitute for IELTS, IDP, the British Council, Cambridge English, ETS or TOEFL. A prep tool that is vague about this is being vague on purpose. You are buying a rehearsal and a measurement, and a trustworthy one says so.

Re-sit the identical paper: the only clean measure of progress

Comparing scores across two different practice tests confounds two things: your improvement and the papers' difficulty. The clean experiment is re-sitting the same paper after focused study — which is why AssessAll sells each versioned mock with three attempts on identical content. Attempt one is your baseline; attempts two and three, taken after you have drilled the weak module, measure movement with the paper held constant. When you want a fresh unseen paper, you buy the next version, not a reshuffle of the last one.

Three attempts on one paper also rehearses the thing nobody drills: the exam's shape. One-play listening, the clock on each module, the long-turn timer — the second time the format costs you nothing, and everything you have is spent on the English.

A four-week readiness plan that ends in one booking

Week 1 — baseline. Take the free Quick Check to place your rough CEFR level, then sit the full four-module mock for your route: the academic paper if you are applying to study, the general-training paper — everyday and workplace listening, practical reading, letter and essay — if you are on a work or migration route to the UK, Canada or Australia. Read the band range and the module scorecards.

Weeks 2–3 — drill the gap, not the syllabus. If listening is the weak module, use accent-specific drills — US, UK and Australian listening are different skills, and the accent you will face on exam day and in the country you are moving to is the one to train. If writing or speaking lags, produce and score more of it; volume with scoring beats another grammar course.

Week 4 — re-measure and decide. Sit attempt two on the identical paper. If the low end of your estimated range now meets your target band with a little margin, book the real exam. If one module still drags, spend attempt three on it a week later. The rule that saves the most money is simple: book when the measurement says so, not when the calendar does.

What it costs

The economics are deliberately lopsided against guessing. AssessAll's English Global mocks start at ₹99 / US$1.99, each with three attempts on the identical paper and AI scoring across all four modules; the free Quick Check costs nothing and places your CEFR level in minutes. The official exam costs the equivalent of US$180–260 per sitting depending on country. A candidate who spends US$2–10 measuring and avoids a single retake has paid for the entire preparation many times over.

Everything runs in a browser on any phone — no test centre, no appointment, no card needed for the free check — which is what makes measured preparation practical from Lagos, Abuja, Nairobi or Accra rather than only from cities with prep schools.

One Skill Retake: the 60-day window that decides when to measure

Since IELTS introduced One Skill Retake, a candidate who misses their target in a single module no longer has to re-sit all four. The rules are narrow and they are what make measuring before you book worth money rather than worth a shrug. Checked at source on 14 September 2026: you may retake one section — Listening, Reading, Writing or Speaking — you must do it within 60 days of the original test, you get one retake per full test, and the original must have been taken on computer at a participating centre. Paper tests are not eligible at all, and UKVI accepts One Skill Retake for Academic and General Training on computer in selected locations only, so a candidate on a UK route has to confirm their centre before they book rather than after.

Read that as a study plan rather than a rule. The 60-day clock starts on exam day, which means the time to find out that writing is your weak module is before you sit, not six weeks after when you are trying to book a retake in a city with no computer-delivered slots. It also means the paper-versus-computer choice is now a strategic one: a paper sitting is a single-shot exam with no partial recovery, and candidates across Nigeria, Kenya, Ghana and South Africa routinely choose a paper centre for reasons of travel or habit without knowing they are giving this up.

The corollary for preparation is the one this guide has argued from the start, now with a deadline attached: a diagnosis is only useful while there is still time to act on it. Sit the mock, drill the module the scorecard names, re-sit the identical paper, and book when the low end of the range clears your target — and book the computer-delivered test, so that a half-band miss in one module is a 60-day problem rather than a whole new exam fee.

What a CEFR level is worth, in the exam board's own words

A lot of advice in this corridor — from agents, from forums, from employers writing job adverts — converts bands to CEFR levels as if the mapping were arithmetic: band 7 is C1, band 5.5 is B2, and so on. IELTS does not agree, and says so on its own CEFR page: “There is not a one-to-one correspondence between IELTS scores and CEFR levels.” The same page describes band 8 as borderline for C2 and says of band 5 that “it is not clear whether band 5 is B1 or B2”. It also tells recognising institutions to set requirements from the band descriptors rather than from a conversion table.

Two practical consequences. First, if a university, employer or route asks for a band, give them a band and do not substitute a CEFR letter you converted yourself; and if they ask for a level, ask which instrument they will accept, because the answer is frequently not IELTS at all. Second, treat any tool that converts your practice score into a confident single letter with the same suspicion this guide asks you to apply to a confident single band — the exam board with the most calibration evidence in the world publishes a range with borderline cases, so a free quiz that publishes a letter is not being more precise, it is being less careful.

If you want to see that argument worked through with the arithmetic, the free workplace English test is eight questions long, scored in your browser, and deliberately refuses to output a CEFR level — with the confidence intervals printed to show why an eight-item test cannot honestly assign one. It is not IELTS preparation and does not pretend to be; it is the clearest short demonstration of what a level claim actually requires.

The real exam publishes its own error, and it is wider than the range this page reports (added 24 September 2026)

This guide argues that an honest mock returns a half-band range rather than a point score, “because that is the honest resolution of a single measurement”. That sentence asserted a quantity and never gave it. IELTS publishes the quantity, on its own test statistics page, and it makes the argument far better than the assertion did — while also going against this site’s own product, which is why it belongs here.

For 2024–2025, IELTS reports a standard error of measurement of 0.43 band for Listening, 0.36 for Academic Reading and 0.47 for General Training Reading, and notes that the SEM “should be interpreted in terms of the final band scores reported for Listening and Reading sections (which are reported in half-bands)”. Reliability on the same page: Listening alphas between 0.83 and 0.95 averaging 0.90, Academic Reading 0.87 to 0.94 averaging 0.92, General Training Reading 0.88 to 0.93 averaging 0.91.

Run the arithmetic a candidate actually needs. One standard error either side of a reported band covers roughly 68 per cent of the plausible range for a true score; the conventional 95 per cent interval is about 1.96 standard errors. On General Training Reading that is about ±0.9 of a band; on Listening about ±0.8; on Academic Reading about ±0.7. A candidate who needs 6.5 and reports 6.5 on Reading has a true score that, on the exam’s own published figures, plausibly sits anywhere from roughly 5.8 to 7.2.

Two consequences follow, and the first one costs this site something. The half-band range on an AssessAll mock report is narrower than the real exam’s own published measurement error on the modules where that error is published. Read it as the resolution of the practice measurement, not as a bound on your exam-day result. Anyone — including us — quoting a tighter range than 0.36 to 0.47 of a band on Reading is quoting a range tighter than the exam publishes for itself.

The second consequence is the one to act on. IELTS is reported in half-bands, and on General Training Reading a single standard error is very nearly one whole reporting step. That means the half-band gap that fails people — a 6.0 where 6.5 was required — sits inside one standard error of measurement. It is real, it is what the receiving institution will act on, and it is also close to the exam’s own noise floor. The practical reading: do not book when your estimate sits at your requirement. Book when it sits clearly above it, because a margin is the only thing that survives measurement error.

And the honest gap at the source. That page publishes an SEM for Listening and Reading and publishes none for Writing or Speaking — for those two it reports inter-rater reliability only, 0.92 for Writing and 0.90 for Speaking. So the two modules that decide most failed attempts, and the two this guide says self-study cannot score, are the two whose measurement error the exam does not publish. That is not a criticism of IELTS; productively rated modules are harder to report an SEM for. It is a reason to hold your estimate on those two modules more loosely still, and to treat a single mock’s writing score as a direction rather than a number.