IELTS Speaking Band Descriptors: Band 6 vs 7 vs 8 Explained
Understand the four IELTS speaking criteria — fluency and coherence, lexical resource, grammar, and pronunciation — and what examiners listen for at each band.
▶ Free College Predictor & study-abroad toolsThe Four IELTS Speaking Criteria
Examiners score speaking on four equally-weighted criteria: Fluency and Coherence (how smoothly and logically you speak), Lexical Resource (vocabulary range and accuracy), Grammatical Range and Accuracy (structures and correctness), and Pronunciation (clarity and natural sound). Each is scored 1–9, and your overall speaking band is their average, rounded to the nearest half-band.
Fluency and Coherence (Band 6 vs 7 vs 8)
Band 6: you speak with some confidence but with occasional hesitation and repetition; links between ideas can feel mechanical. Band 7: you speak mostly fluently with only occasional hesitation, linking and developing ideas logically, and can sustain discussion of complex topics. Band 8: you speak fluently and naturally with minimal hesitation, using sophisticated discourse markers and handling complex discussion without losing coherence.
Lexical Resource (Vocabulary Range)
Band 6: appropriate vocabulary but limited range, with noticeable repetition and some inaccurate idiomatic use. Band 7: a flexible range of vocabulary used accurately, including some less common words, with only minor errors in idioms or collocations. Band 8: a wide range used naturally and precisely, with confident, accurate use of idiomatic language adapted to context.
Grammatical Range and Accuracy
Band 6: a mix of simple and complex structures but with frequent errors and some inconsistency in tenses. Band 7: a variety of complex structures used accurately, with errors that are rare and usually in less common structures. Band 8: a wide range of structures used accurately and appropriately, with only rare, minor errors. Even strong speakers make occasional slips — perfection is not required.
Pronunciation: Clarity and Accent
Band 6: generally clear despite a noticeable accent; some sounds may be mispronounced and word stress can occasionally affect meaning. Band 7: clear pronunciation with mostly appropriate stress and intonation and only occasional lapses. Band 8: clear and natural pronunciation with effective stress, rhythm and intonation, and only very rare mispronunciations.
Strategy to Improve Your Speaking Band
Identify your weakest criterion — often fluency or vocabulary — and focus there. For fluency, speak aloud daily and record yourself; for vocabulary, learn chunks and collocations, not isolated words; for grammar, review your error patterns; for pronunciation, mirror native speakers. Practise full mock speaking tests and self-assess against the descriptors. Practise free IELTS speaking on LandingPrep to track development across all four criteria.
The four criteria are weighted equally, and that changes what to practise
Your Speaking band is the average of Fluency and Coherence, Lexical Resource, Grammatical Range and Accuracy, and Pronunciation — each worth exactly a quarter. That arithmetic has a practical consequence most candidates miss: your weakest criterion drags the whole band, and improving something you are already good at is close to worthless. A candidate at 7/7/7/5 averages 6.5, which is reported as 6.5. Lifting that 5 to a 6 takes the average to 6.75, reported as 7.0 — half a band from one criterion. The arithmetic is the same wherever the half band comes from, so why start with the weakest? Two reasons that have nothing to do with the maths. Going from 5 to 6 is almost always a shorter distance than going from 7 to 8, because the lower descriptors describe basic gaps you can close deliberately while the higher ones describe range and naturalness that take much longer. And where an institution sets a minimum in EACH criterion, a weak one disqualifies you no matter what the average says.
Fluency is not speed, and hesitation is not automatically penalised
The descriptors distinguish between hesitating to search for language and hesitating to think about content. Pausing because you are considering an idea is normal speech and is not marked down; pausing because you cannot find the word is. This is why speaking faster is the wrong fix and usually makes things worse — it produces more self-correction and more abandoned sentences, both of which the descriptors do penalise. What actually raises Fluency is being able to keep going: extending an answer with a reason and an example, and paraphrasing around a word you cannot retrieve rather than stopping to hunt for it.
Lexical Resource rewards precision, and punishes rehearsed phrases
The higher bands ask for flexible, precise vocabulary used naturally — including idiomatic language. What they do not reward is a memorised list deployed regardless of context. Examiners assess thousands of candidates and recognise inserted phrases immediately; language that is clearly pre-learned and does not fit the question is discounted rather than credited. The skill that genuinely raises this criterion is paraphrase: being able to explain a word you cannot recall demonstrates exactly the flexibility the descriptors describe, while going silent demonstrates its absence.
Grammatical Range and Accuracy is a trade-off, not two separate scores
The descriptors assess both the variety of structures you attempt and how much of your speech is error-free. These pull against each other, and candidates fail at both ends. Speaking only in short simple sentences caps your range no matter how accurate it is. Attempting complex structures that collapse mid-sentence caps your accuracy. The band comes from the proportion of your speech that is error-free while still showing variety, which means controlled complexity beats ambition — use the structures you can land reliably, and add one you are working on rather than several you are not.
Pronunciation assesses intelligibility, and an accent is not a penalty
This is the most misunderstood criterion. The descriptors are about whether you can be understood without strain, and they explicitly do not require or reward any particular accent. What is assessed is the features that carry meaning: individual sounds, but more importantly word and sentence stress, rhythm and intonation. Stress placement usually affects intelligibility more than individual sounds do — a sentence with every sound approximately right but the stress in the wrong place is harder to follow than the reverse. Candidates who spend months trying to sound British or American are working on the thing that is not being measured.
How to use the descriptors on your own practice
Record a full two-minute Part 2 answer in one take, then score it four times, once per criterion, listening for a different thing each pass. This is slow and it is the only method that tells you where your band actually is. Scoring an answer once as an overall impression gives you a number you cannot act on. Four passes give you a specific target — and since a single criterion is a quarter of your result, that target is worth roughly half a band on its own. Then re-record the same cue card addressing only that criterion, and compare.
Keep going — free practice
Last updated: