IELTS Writing Mock Test: How to Run One Properly, and What to Measure Afterwards

Most candidates take a writing mock test and learn one number from it. That number is the least useful thing the paper can tell you. The hour is the easy half; what you do with the two answers afterwards is what moves a band.

This post gives you a complete mock paper to sit, the exact conditions that make it worth sitting, and a marking routine for afterwards. It also shows, with a count you can check, why one overall band from a mock test hides what you most needed to see.

A mock test is not practice

These are two different activities and mixing them wastes both.

Practice is slow. You write one paragraph, you look up a word, you check whether your grammar is right, you rewrite it. The goal is to change something you do.

A mock test is a measurement. One hour, no help, no stopping, both tasks, however badly it goes. The goal is to find out what you produce under pressure, which is almost never what you produce with time to think.

If you practise all the time and never test, you will be surprised on exam day by how little you can produce in 40 minutes. If you test all the time and never practise, you repeat the same errors at the same speed for months. Most candidates who feel stuck are testing without marking, which is the most common way to work hard and improve nothing.

A reasonable rhythm is one mock test a week and practice on the other days. More often than that and you have no time to fix anything between tests.

The conditions that make it a real test

The paper is 60 minutes for both tasks. No extra reading time. The official instruction is to spend about 20 minutes on Task 1 and about 40 minutes on Task 2, and to write at least 150 words for Task 1 and at least 250 words for Task 2. Task 2 carries more weight in your writing band than Task 1 does, which is why it gets double the time.

Set these up before you start the clock.

• One timer for 60 minutes, not two. In the real test nobody tells you when 20 minutes have passed. Managing that split is part of what is being measured, and it is where many candidates lose Task 2. Our guide to time management in IELTS writing has the minute by minute plan.
• Write the way you will write on the day. If you are sitting the paper based test, write by hand on lined paper. If you are sitting the computer based test, type in a plain text editor with spellcheck turned off. These produce different word counts and different error rates, so a typed mock does not predict a handwritten exam. Handwriting in IELTS covers what is readable enough to mark.
• No dictionary, no grammar checker, no searching the topic. Not even once. One lookup makes the whole paper unmarkable as a measurement.
• Do not stop when the time is up. Draw a line at the 60 minute mark and then, if you want, keep writing in a different colour. The part after the line is not part of the test, but it tells you how close you were.
• Count your words at the end, by hand. Count the words in three lines, take the average, multiply by your number of lines. You need to know this number before the exam, because you will not be able to count 250 words in the real thing.

A mock paper you can sit now

Both questions below are written in the style of the exam. They are not real past papers, because real papers are not published, but they are the same shape and the same difficulty.

Task 1, Academic. Spend about 20 minutes. Write at least 150 words.

The table below shows the number of hours per week that people in four age groups spent on paid work, housework and leisure in one country in 2024. Summarise the information by selecting and reporting the main features, and make comparisons where relevant.

All figures are hours per week.

Task 1, General Training. Use this instead if you are sitting General Training. Spend about 20 minutes. Write at least 150 words.

You recently attended a training course that your employer paid for. Write a letter to your manager. In your letter, say what you learned on the course, explain how you will use it in your job, and suggest which colleagues should attend it next.

Task 2, both versions. Spend about 40 minutes. Write at least 250 words.

Some people believe that governments should spend money on public transport rather than on building new roads. Others think new roads are the better investment. Discuss both views and give your own opinion.

That is a complete paper. Stop reading, sit it, and come back. The rest of this post is about what to do with the two answers in front of you.

Marking it: four criteria, separately

Your writing band comes from four criteria, each marked on its own, each worth the same amount. For Task 2 they are Task Response (did you answer the whole question), Coherence and Cohesion (organisation and linking), Lexical Resource (vocabulary) and Grammatical Range and Accuracy (sentence variety and errors). For Task 1 the first one is called Task Achievement instead, and it asks whether you selected and summarised the right information.

The two tasks are not averaged equally. Task 2 counts for roughly two thirds of your writing band and Task 1 for roughly one third, and the result is then rounded. Exactly how that arithmetic works is in how your writing band score is calculated, and the score calculator shows what a given set of four criteria produces.

Mark each answer against each criterion separately, in four passes. Four passes sounds slow and is the only way to see anything. Reading your essay once and feeling that it is "about a 6" tells you nothing you can act on.

Pass one, Task Response. Find the question again. Underline every part of it. Does your answer deal with every part? An essay that discusses both views but never states your own opinion has not answered a "give your own opinion" question, however well written it is.

Pass two, Coherence and Cohesion. Read only your first sentence of each paragraph. Do those sentences, read together, make a sensible summary of your argument? If they do not, your paragraphs are not organised around one idea each.

Pass three, Lexical Resource. Circle every word you repeated more than twice. Mark every word you were not completely sure about.

Pass four, Grammar. Count your errors, and count your sentence types. How many complex sentences did you write? If every sentence has the same shape, the range half of this criterion is capped whatever your accuracy is.

Why one overall band is the wrong thing to take from a mock test

Here is the count that changed how I would use a mock paper.

The Bandly sample library publishes 25 full answers, each with a band estimate for all four criteria on the page. I read every record and took the difference between the highest and the lowest of those four estimates in each answer. The result:

Twenty one of the 25 answers have a gap of at least two bands. The mean gap is exactly two bands. Lexical Resource was the lowest or joint lowest estimate in 23 of the 25.

That is the argument against taking one number from a mock test. An answer with estimates of 8, 6, 6 and 7 and an answer with estimates of 7, 7, 7, and 6 can round to the same overall band while needing completely different work. The first candidate needs vocabulary and organisation. The second needs grammar. One number cannot tell them apart.

Two honest limits on that count. These are estimates produced by Bandly's own engine, and no qualified examiner has checked them, so the bands themselves are not evidence of anything. And the highest criterion was Task Response or Task Achievement in all 25, which is partly a known tendency of our engine rather than a fact about candidates. The useful part is the size of the gap, which is large and consistent, and matches what the written feedback on those pages actually complains about.

Where free mock tests with a score fit in

Searching for a writing mock test online with a free score gets you three kinds of result, and they are not equally useful.

An automated estimate is most useful for the comparison rather than the number: run the same essay and your rewrite of it through the same tool and see which criterion moved. Bandly's own estimates have never been compared against a qualified examiner, so no automated tool on this page, including ours, should be treated as an exam result.

The cheapest honest check is to compare your answer against a full answer at the level you want. The sample library publishes complete answers with every scored sentence marked and the reasoning next to it, including Task 1 Academic reports for the table question above. Read the sentences rather than the bands.

What to do today

1. Sit the paper above, in one 60 minute block, under the conditions in this post.
2. Mark it in four separate passes, one per criterion, and write down a band for each.
3. Take your lowest criterion and nothing else. Spend the week on that one.
4. Next week, sit a different paper and see whether that one criterion moved.

If you want a per criterion read on the two answers you just wrote, you can paste each one into Bandly and get a band estimate for all four criteria with the sentences that drove each one marked on the page. Use it for the shape of the gap, not as a score.