IELTS Writing Task 2: the essay that decides your band

Task 2 is worth twice as much as Task 1, and it’s marked on four criteria you can plan for. This guide covers the five question types, a structure that survives exam pressure, a full model answer with examiner commentary, and what actually separates Band 6 from Band 7.

What Task 2 is — and why it decides your Writing band

Task 2 is a 250-word (minimum) discursive essay on a general topic: education, technology, environment, society, health. You get a position or a question, and 40 of your 60 Writing minutes to deal with it.

Here is the part many candidates never hear: Task 2 counts twice as much as Task 1 toward your Writing band. If you score 6.0 on Task 1 and 7.0 on Task 2, you're on roughly 6.5. Flip those scores and you're on 6.0. When preparation time is short, Task 2 is where the marks are.

How it's scored, in plain language

Four criteria, 25% each. Examiners score each one separately, then average. You don't need perfection anywhere — you need no weak corner.

Notice what is not on the list: long words, a fixed number of paragraphs, "advanced" idioms, or a memorised opinion. Examiners read hundreds of essays; templates announce themselves within two lines.

The five question types (and what each one demands)

Every Task 2 question is one of five patterns. Identify the pattern before you plan — each has a different definition of "fully answered".

• Opinion (agree/disagree). "To what extent do you agree?" — State a clear position in the introduction and hold it to the end. A wobbling position is the classic Task Response killer.
• Discussion. "Discuss both views and give your own opinion." — You must do all three. Skipping your own view, or one of the views, caps you at 5.0–6.0 for Task Response no matter how good the language is.
• Advantages / disadvantages. Sometimes "do the advantages outweigh…?" — if the question says outweigh, the examiner expects a verdict, not two balanced lists.
• Problem / solution. "What problems does this cause and how can they be solved?" — Link each solution to a problem you named. Solutions to problems you never mentioned read as off-topic.
• Two-part (direct questions). Two explicit questions in one prompt. Answer both, with roughly equal weight — one body paragraph each is the cleanest split.

A structure that holds under pressure

Four paragraphs. Not because four is magic, but because it forces one idea per paragraph — which is what Coherence & Cohesion rewards.

1. Introduction (2–3 sentences). Paraphrase the topic in your own words, then answer the question directly. No hook, no "in the modern era", no essay-about-the-essay.
2. Body 1 (one idea, developed). Topic sentence → explain → example or consequence → link back to the question.
3. Body 2 (second idea, same treatment). If the task is a discussion, this is the other view — plus a sentence showing where you stand.
4. Conclusion (1–2 sentences). Restate your position in fresh words. Never introduce a new idea here.

One developed example beats three listed reasons. "For example, when my city introduced congestion charging in 2022, traffic fell but small shops near the boundary struggled" develops an argument. "There are many examples of this in many countries" develops nothing.

A Band 8+ model answer, taken apart

The seven mistakes examiners see every day

• Answering the topic, not the question. The prompt asks about university admissions; the essay argues education is important. Band 5 territory for Task Response, regardless of grammar.
• The memorised introduction. "In this contemporary epoch, it is an incontrovertible fact that…" — examiners are trained to spot rote language, and it drags Lexical Resource down, not up.
• Linker wallpaper. Moreover / Furthermore / In addition at the start of every sentence signals mechanical cohesion — the opposite of what CC rewards at 7+.
• No error-free sentences. The 7.0 grammar descriptor expects frequent error-free sentences. Three clean sentences beat ten ambitious broken ones.
• Under-length or over-planned. Below 250 words caps Task Response. But 400+ words usually means less checking time and more errors — 270–320 is the working range.
• Examples from nowhere. "Studies show" with no study. Use plausible, specific experience instead — examiners score development, not citations.
• Conclusion introduces a new idea. It reads as an unplanned essay. Conclusions restate; they never add.

A practice method that moves the band

Writing more essays is not the method. Writing, getting criterion-level feedback, and rewriting is. One essay processed properly is worth five essays filed away.

1. Session 1 — plan only (15 min). Take three past questions. For each: identify the type, underline the jobs, write a four-line plan. Don't write the essay. Planning is the highest-leverage skill in Task 2 and almost nobody practises it in isolation.
2. Session 2 — timed essay (40 min). One question, full exam conditions, 40 minutes including 3 minutes of checking. Checking targets: articles, subject–verb agreement, singular/plural — the three highest-frequency error families.
3. Session 3 — feedback + rewrite (30 min). Get the essay scored per criterion, find your weakest of the four, and rewrite only the paragraph that failed it. This is where the band actually moves.
4. Repeat across the five question types — most candidates discover they're fine on opinion essays and consistently underperform on two-part questions, or the reverse.

From 6.0 to 7.0: what actually changes

The jump from 6 to 7 is not "better vocabulary". Reading the descriptors side by side, three things separate the bands:

Practical translation: to move up a band, cut one idea per essay and develop the survivors; hunt down your three commonest grammar errors and eliminate them; and replace impressive-sounding words with exact ones. That is a four-to-eight-week project at one focused essay every other day — not a weekend one.