The IELTS fee is the most expensive quiz entry on earth if you walk in not knowing your level.
Yet your band score is one of the most predictable numbers in education: two sections convert directly from raw scores on published tables, and the other two can be measured against public criteria. Prediction is not fortune telling - it is arithmetic plus honest practice.
Short answer: predict your IELTS band by averaging your last three timed practice scores per section - Listening and Reading convert from raw scores on the official tables, Writing needs a criterion-level check rather than a guess, and Speaking needs recorded practice rated against the four descriptors.
Done honestly, prediction lands within half a band of the real result.
This guide covers the conversion numbers, the method for the two "unpredictable" sections, the noise that ruins predictions, and the booking rule that turns your prediction into money saved.
Predicting Listening and Reading: pure arithmetic
Both sections have 40 questions and no partial credit, so a raw score maps to a band on a conversion table. The approximate thresholds as of 2026:
| Band | Listening (out of 40) | Academic Reading | General Training Reading |
|---|---|---|---|
| 6.0 | 23 | 23 | 30 |
| 6.5 | 26 | 27 | 32 |
| 7.0 | 30 | 30 | 34 |
| 7.5 | 32 | 33 | 36 |
| 8.0 | 35 | 35 | 38 |
Exact cut-offs shift by a question or two between test versions, which is why every serious source calls these approximate.
Note the General Training column: GT Reading demands about four more correct answers for the same band, and predictors that ignore the difference mislead GT candidates by a full band. Our conversion table guide has the full band-by-band breakdown.
To predict: sit a full timed section, count correct answers, convert. The free Band Score Calculator does the conversion for both modules and combines the four sections into a projected overall with the official rounding rules applied.
Predicting Writing: replace guessing with measurement
Writing has no raw score - examiners grade Task Response, Coherence and Cohesion, Lexical Resource, and Grammatical Range and Accuracy, each worth 25%. That is why Writing is the section candidates mispredict most, usually optimistically: self-assessment against criteria you have not internalised is guessing with extra steps.
The fix is a criterion-level check on timed essays. The IELTSbiz Writing Checker returns a band per criterion plus sentence-level evidence, and the estimate is validated: IELTSbiz's study of 1,200 essays found AI band scores matched human examiners within ±0.5 bands in 94.2% of cases.
For prediction purposes, check three timed essays on different question types and take the average. A single essay predicts one performance; three predict your level.
Predicting Speaking: record, then rate
Speaking is graded on Fluency and Coherence, Lexical Resource, Grammatical Range and Accuracy, and Pronunciation. Without an examiner in the room, your best predictor is recorded practice rated against those descriptors - by a teacher, a study partner, or yourself after studying what each band sounds like.
The method: record a full mock interview - Part 1 questions, a Part 2 cue card with one minute of preparation, Part 3 follow-ups - then listen back against the public descriptors.
Count long pauses and self-corrections for Fluency; notice repeated words for Lexical Resource; listen for whether complex sentences survive to their end for Grammar. Our guide to how Speaking is scored decodes each criterion.
Self-rated Speaking predictions run optimistic by about half a band on average, so if your self-assessment says 7.0, plan around 6.5 unless a teacher confirms otherwise.
The three input errors that break predictions
Error one: untimed practice. Reading scores collected without the 60-minute clock overpredict by half a band or more, because time pressure - not comprehension - is what fails most candidates. Every practice that feeds your prediction must be timed.
Error two: familiar material. Rescoring a test you half-remember, or practising from the same book for the third time, measures memory rather than skill. Fresh material only - generated Cambridge-style passages are useful here precisely because you cannot have seen them before.
Error three: single-test predictions. Any one practice carries noise - topic luck, tiredness, one bad audio section - of up to half a band. Your level is the average of your last three timed practices per section, and decisions as expensive as a test booking deserve the averaged number.
From four predictions to one overall band
The overall band is the mean of the four section bands rounded to the nearest half band, and the rounding favours you: a mean ending in .25 rounds up to .5, and .75 rounds up to the next whole band. Listening 7.0, Reading 6.5, Writing 6.0, Speaking 6.5 averages 6.5 exactly.
Model both the overall and the sections, because most requirements specify minimums per section: a 7.0 overall with a 5.5 in Writing fails many university and visa thresholds that a flat 6.5 would pass. Our guides to how band scores work and what counts as a good score map the common requirement patterns.
The calculator lets you play the useful "what if" games: if Writing stays at 6.0, what do the other three sections need to average? Usually the answer reveals that fixing the weak section is cheaper than perfecting the strong ones.
Improving the prediction: track by question type
A prediction tells you where you are; a plan needs to know why. In Reading and Listening, the why lives at question-type level: two candidates both scoring 30/40 might lose their ten marks to completely different formats - one to True/False/Not Given, the other to matching headings.
This is where per-question-type tracking earns its keep. IELTSbiz reading practice scores every session by question type and tracks an estimated band per type over time, so your prediction comes with its own repair manual: the exact formats leaking marks, ranked.
Five recovered answers in your weakest type is the difference between 7.0 and 7.5-8.0 territory.
Free practice covers daily sessions; unlimited practice and tracking is $6.99/month or $29 lifetime (pricing).
The booking rule: when the prediction says go
Set the threshold before you look at the data: book the test when your three-practice average meets or exceeds your target in every section for two consecutive weeks. Not one great day - two stable weeks.
If the average sits half a band below target, a focused four-to-six-week block on the weakest section is usually enough, and it costs a fraction of a retake.
If the gap is a full band or more, adjust the timeline honestly - our guide to how long IELTS preparation takes gives realistic timelines by starting level.
Candidates who predict and then book beat candidates who book and then hope, for the least glamorous reason in test preparation: they only pay the fee once.