In one line: Band descriptors are IELTS’s official rubric — the exact criteria examiners grade against. Most aspirants prepare blind to them; knowing the 7.0 line for each skill tells you precisely what to fix.

In fact, band descriptors are the exam’s rubric — the exact criteria examiners grade against. Most aspirants prepare blind to them; knowing the 7.0 line for each skill tells you precisely what to fix. This note translates the four skills’ descriptors into plain, actionable terms.

Quick Answer: A 7.0 is not “very good English” — it is a checklist. Knowing the 7.0 line for each skill tells you precisely what to fix. This note translates the four skills’ band descriptors into plain, actionable terms, so your practice becomes criterion-targeted repair instead of generic study.

Table of Contents

1. How Scoring Works
2. Writing Task 2: The Four Criteria at Band 7
3. Speaking: The Four Criteria at Band 7
4. Reading and Listening: The Raw-Score Ladder
5. Diagnosing Your Gap
6. How Exams Probe This Topic
7. Quick Revision: One-Glance Facts

1. How Scoring Works

• Each skill is scored 0–9 in whole and half bands. Therefore, the overall band = the average of the four skills rounded to the nearest half band (with .25 rounding up: 6.5 + 6.5 + 6 + 6 = 6.25 → 6.5 overall). Writing and Speaking are graded against published band descriptors by trained, certified examiners. Meanwhile, Reading and Listening are raw-score conversions — you simply need a set number of correct answers out of 40.
• Why descriptors matter. Writing 6.0 → 6.5 is the most common overall-band blocker — and the fix depends entirely on which criterion (Task Response, Coherence & Cohesion, Lexical Resource, Grammatical Range & Accuracy) is bleeding the marks. Preparing without knowing which criterion is weak is like studying for an exam without knowing the syllabus: you may work hard on the wrong things entirely.

2. Writing Task 2: The Four Criteria at Band 7

Indeed, here is a plain-language translation of the public descriptors, with the tell-tale symptoms of band 6 included so you can self-diagnose:

1. Task Response 7: position clear throughout; main ideas developed and extended (not just listed); all parts of the question fully addressed. The 6-vs-7 gap is usually “ideas present but not extended” — one developed paragraph beats three shallow ones. At band 7, the examiner wants to see you take an idea and push it: give the reason, the consequence, the example.
2. Coherence & Cohesion 7: logical progression throughout; a clear central topic in each paragraph; cohesive devices used flexibly — not mechanically. The band-6 symptom: “Firstly, Secondly, Moreover” bolted onto every sentence. At band 7, links should feel invisible: referencing (“this policy,” “such measures”), substitution, and clear paragraph logic do most of the work.
3. Lexical Resource 7: sufficient range for nuance; some awareness of style and collocation; occasional errors only. The band-6 symptom: repeated topic words (“education… education… education”) instead of referencing and variation. Collocation awareness matters more than rare vocabulary — “impose a ban” scores higher than “make a prohibition.”
4. Grammatical Range & Accuracy 7: a variety of complex structures; frequent error-free sentences; errors rare and non-impeding. The key line is “frequent error-free sentences,” meaning roughly 60%+ of your sentences are clean. Range without accuracy is band 6; accuracy without range is also band 6 — you need both.

3. Speaking: The Four Criteria at Band 7

1. Fluency & Coherence 7: speaks at length without noticeable effort; uses connectives and discourse markers naturally. The Part 5 AREC method (Answer, Reason, Example, Consequence) displays exactly this — extended, organised turns rather than one-sentence answers followed by silence.
2. Lexical Resource 7: uses vocabulary flexibly across a range of topics; paraphrases effectively; uses some idiomatic items. Paraphrasing the question is the examiner’s first cue: if the question says “should children learn a second language” and you open with “whether youngsters should acquire a foreign tongue,” you have shown the skill immediately.
3. Grammatical Range & Accuracy 7: a mix of simple and complex structures with frequent error-free stretches; complex forms (conditionals, relative clauses) appear even if not flawless. The examiner is listening for attempts at complexity, not perfection — a slightly imperfect conditional still earns range credit; avoiding conditionals entirely does not.
4. Pronunciation 7: easy to follow throughout; intonation and stress used with mostly effective control. Accent is irrelevant — intelligibility and range of features are the criteria. A strong regional accent that never causes misunderstanding is fully compatible with band 9.

4. Reading and Listening: The Raw-Score Ladder

Approximate raw scores for band 7 (Academic; varies slightly by test form):

• Listening: ~30–32 of 40 correct for 7.0 (35+ for 8.0). That is a maximum of 8–10 wrong answers across four sections.
• Academic Reading: ~30–32 of 40 for 7.0 (35+ for 8.0). General Training Reading needs ~34–35 for 7.0 — the texts are easier, so the conversion is tighter.
• The planning use: convert your target band into a raw target per section and per question type. Your error log tells you which type — True/False/Not Given, matching headings, multiple choice — costs you the marks. A candidate losing four questions to TFNG does not need “more reading practice”; they need fifteen minutes of TFNG technique work.

5. Diagnosing Your Gap

• The mock autopsy. One evaluated mock (an examiner-checked or AI-checked essay, plus a recorded speaking test) with criterion-level feedback beats ten unchecked mocks. The fix is criterion-specific: a cohesion-6 needs linking and paragraphing drills, not vocabulary lists. Without criterion-level feedback, you cannot tell whether your writing is bleeding marks in Task Response or Grammar — and the remedies are completely different.
• The half-band project. Writing 6 → 6.5 typically needs: full question coverage, one extended idea per paragraph, varied linking, and error-proofing the tense–article–subject-verb trio — a four-item checklist, not “better English.” This is deliberate, narrow work measured over two or three essays, not months of general practice.
• The score-protection move. If Listening is your 8 and Writing your 6, the overall is capped by Writing. Allocate practice hours by the descriptor gap, not by preference. Candidates naturally drill their strongest skill because it feels good — resist this. The half band lives in your weakest criterion.

6. How Exams Probe This Topic

• The official source: the publicly available band descriptors on the IELTS official website — read them once verbatim. This note is their plain-language operating version, but the originals are only a few pages long and worth seeing first-hand.
• Cross-reference: Part 8 of this series (the 8-week cycle) schedules descriptor-based checkpoints — the week-4 and week-6 evaluations. Those checkpoints are where the diagnosis-and-repair method in this note gets applied on a calendar.

7. Quick Revision: One-Glance Facts

• The four writing criteria: Task Response, Coherence & Cohesion, Lexical Resource, Grammatical Range & Accuracy — at 7: developed ideas, flexible cohesion, nuanced vocabulary, frequent error-free sentences.
• The four speaking criteria: Fluency & Coherence, Lexical Resource, Grammatical Range & Accuracy, Pronunciation — at 7: length without effort, flexible vocabulary, mixed structures, easy-to-follow speech.
• Raw ladder: Listening and Reading ~30–32/40 for 7.0 (General Training Reading ~34–35).
• Rounding: .25 rounds up — two 6.5s plus two 6.0s = 6.5 overall.
• The rule: diagnose by criterion, fix by criterion.

A 7.0 is not “very good English” — it is a checklist: developed ideas, flexible cohesion, paraphrase, error-free frequency, sustained speech. Read the descriptors once, diagnose your mock against them, and your preparation stops being generic practice and becomes criterion-targeted repair — the fastest route through the half-band ceiling.

Practice Corner: Five More Checks (with Answers)

1. Writing Task 2 is assessed on which four criteria? — Task Response, Coherence & Cohesion, Lexical Resource, and Grammatical Range & Accuracy.
2. Fluency band 7 requires what? — Speaking at length without noticeable effort, with natural connectives and discourse markers.
3. Listening at roughly 30–32/40 gives which band? — 7.0 (35+ of 40 for 8.0).
4. The overall score for 6.5 + 6.5 + 6 + 6 rounds to? — 6.5 (the .25-rounds-up rule: the average of 6.25 rounds up to 6.5).
5. Pronunciation band 7 means what? — Easy to follow throughout; accent is irrelevant — intelligibility and feature range are the criteria.

The Self-Marking Protocol (When No Teacher Is Available)

Mark your own Task 2 essay in three passes, one criterion at a time:

Pass 1 — Task Response: underline your position in the introduction, then check that every body paragraph develops it rather than merely restating it.

Pass 2 — Coherence: count your linking devices and check each paragraph has exactly one central topic. If a paragraph serves two ideas, split it or cut one.

Pass 3 — Language: highlight the first error in every sentence, then classify the highlighted errors by type (articles, tense, agreement, collocation). The classification tells you exactly which grammar chapter to reopen — most candidates find 60% of their errors come from just two categories.

This protocol converts an unmarked mock into a criterion-targeted repair plan — which is the entire point of practising writing at all.

The Band-7 Boundary Cases (The Marker’s Hardest Calls)

The descriptors’ edge questions, resolved:

• “A variety of complex structures” — the marker wants conditionals, relative clauses, and passive transformations actually used, not just known. One accurate complex sentence per paragraph satisfies variety better than four tangled ones.
• “Frequent error-free sentences” — the working interpretation is roughly sixty percent clean sentences. This means the candidate with simple but clean grammar outranks the one with ambitious but broken complexity: accuracy precedes ambition at this band.
• “Flexibly” (lexical) — paraphrase demonstrated, not vocabulary rarity. A candidate who writes “young people,” then “the younger generation,” then “minors” across an essay shows flexibility no memorised polysyllable can.
• Coherence 7’s “logical progression” — the paragraph test: delete your linking words. If the logic still reads, the coherence is real; if it collapses, it was glue holding nothing.

Four boundary cases, four working rules — the marks at the margin go to accuracy, paraphrase, and structure in itself, in that order.

Read next: IELTS Prep Part 8: The 8-Week Practice-Test Cycle

Related Exam Guides

• IELTS Prep Part 8: The 8-Week Practice-Test Cycle
• IELTS Prep Part 7: Topic Vocabulary — Word Lists and the Collocation Habit
• IELTS Prep Part 5: Speaking Part 3 — The Follow-Up Depth Method

Frequently Asked Questions

What should you know about how IELTS scoring works?

Each skill is scored 0–9 in whole and half bands. The overall band = the average of the four skills rounded to the nearest half (with .25 rounding up: 6.5 + 6.5 + 6 + 6 = 6.25 → 6.5). Writing and Speaking are graded against published band descriptors by trained examiners; Reading and Listening are raw-score conversions.

What should you know about the four Writing Task 2 criteria at band 7?

Plain-language translation of the public descriptors: Task Response 7 requires a position clear throughout, main ideas developed and extended (not just listed), and all parts of the question fully addressed. The 6-vs-7 gap is usually “ideas present but not extended” — one developed paragraph beats three shallow ones.

What should you know about the four Speaking criteria at band 7?

Fluency & Coherence 7: speaks at length without noticeable effort; uses connectives and discourse markers naturally — the Part 5 AREC method displays exactly this. Lexical Resource 7: uses vocabulary flexibly for a range of topics; paraphrases effectively; some idiomatic items — paraphrasing the question is the examiner’s first cue.

What should you know about the Reading and Listening raw-score ladder?

Approximate raw scores for band 7 (varies slightly by test form): Listening ~30–32 of 40 for 7.0 (35+ for 8.0); Academic Reading ~30–32 of 40 for 7.0 (35+ for 8.0); General Training Reading needs ~34–35 for 7.0 (easier texts, tighter conversion). The planning use: convert your target band to a raw target per section and per question type — your error log tells you which type (True/False/Not Given, matching headings) costs the marks.

What should you know about diagnosing your gap?

The mock autopsy: one evaluated mock (an examiner-checked or AI-checked essay, plus a recorded speaking test) with criterion-level feedback beats ten unchecked mocks — the fix is criterion-specific; a cohesion-6 needs linking and paragraphing drills, not vocabulary lists. The half-band project: Writing 6 → 6.5 typically needs full question coverage, one extended idea per paragraph, varied linking, and error-proofing the tense–article–subject-verb trio — a four-item checklist, not “better English.”

Is my accent a problem for the Speaking band score?

No. Accent is explicitly irrelevant to the Pronunciation criterion. What matters is intelligibility — can the examiner follow you easily throughout — and whether you use features such as intonation, sentence stress, and chunking with mostly effective control.

How long does it take to move from Writing 6.0 to 6.5?

There is no fixed timeline, but candidates who work criterion-by-criterion — using evaluated mocks and a targeted checklist (question coverage, extended ideas, varied linking, tense-article-agreement proofing) — typically move faster than those practising generically, because every hour of practice addresses a documented gap.

Related explainers

Exam Notification Digest: Government Exams This Fortnight — Dates, Vacancies, Eligibility

An IELTS Speaking Mock: Full Transcript with Band Fixes

Part 5: Cause, Effect and Hypothesis Patterns in IELTS Speaking