Am I Ready for IELTS? Five Checks That Beat a Gut Feeling

You want someone to look at your Cambridge 18 results and say yes, book it. Nobody honest will do that. The friend who tells you "you'll get 7 easily" has never marked a script against a band descriptor, and the coaching centre that wants your enrolment fee is the last place to ask.

There's one real answer to am I ready for IELTS, and it's a number. Not a feeling, not how many books you've finished — a band produced under conditions close enough to the real sitting that the band means something. So the question worth asking is: what makes a practice band trustworthy enough to bet a non-refundable fee on?

Five checks below. Three are about the test. Two are about the measuring instrument, and those two are the ones almost nobody applies.

"Am I ready for IELTS?" is the wrong question

Your sense of readiness is assembled from your best evenings. You remember the passage on urban beekeeping where True/False/Not Given finally clicked and you got 13 out of 14. You don't average that against the four sessions where you ran out of time in passage 3 and guessed six answers off the back of a heading.

The cost of being wrong is lopsided, too. Cancel late and the centre keeps most of the fee; the rules on transfers and refunds belong to them, not you, and they aren't generous. The slot you actually want — a weekend, a venue you can reach across Dhaka without leaving at six in the morning — fills first. Computer-delivered results come back in about 3-5 days, paper-based in around 13. That gap alone can decide whether a retake fits before your university deadline.

So the question needs an instrument. Fail any of the five checks and the answer is "not yet", regardless of how the last mock felt.

Check 1 — one sitting, three papers, nothing in between

Most candidates have never done this even once. They've done Listening on a Tuesday, two Reading passages on Wednesday, a Task 2 on Saturday afternoon with the phone face-up next to the notebook. The real sitting is Listening, then Reading, then Writing, with no break you'd call a break: roughly two and a half hours of unbroken English, and the last thing you do, when your concentration is at its worst, is the 250-word essay that carries twice the marks of Task 1.

Computer-delivered IELTS arrived in Bangladesh later than the paper version, so a lot of people in Dhaka have trained entirely on paper and will sit on a screen at a British Council or IDP centre. Those aren't the same test. You type answers straight into the box, so spelling counts as you go with nothing to tidy up afterwards; there's no ten minutes at the end to catch a missing plural. Reading is split-screen with a highlight tool on right-click. Part 4 of Listening runs as one continuous monologue.

Pass condition for Check 1: one complete Listening, Reading and Writing block, timed, in exam order, on the format you're actually booked to sit. If you've never done that, you don't have a band. You have an impression.

Check 2 — one band is a data point, two is a level

Sit a second full paper about a week after the first, from a different Cambridge book. Then compare.

Two things move a single score, and neither has anything to do with your English. The first is version variance in what you happened to get: a Reading passage about textile manufacturing lands differently for a Bangladeshi candidate than one about Antarctic ice cores, and the raw-to-band conversion adjusts for how hard the paper was overall, not for what you personally happen to know. The second is you — how you slept, whether the construction next door started at seven, whether you sat down already annoyed.

Two sittings a week apart give you a range instead of a point. If the two bands are within half a band of each other, you have a level and you can plan around the lower one. If they're a full band apart, wait. A gap that size is variance, and booking on the high one is wishful.

Here's where I'll cut against the standard advice: doing more mock tests does not raise your score, and in the last fortnight it usually lowers it. A mock is a thermometer. Taking your temperature eight times in ten days doesn't treat the fever, and it burns the Cambridge papers you needed for a clean measurement later. Two full sittings a fortnight is plenty. The hours in between are where the work happens.

Check 3 — sit half a band above what you actually need

Your per-section scores move between test versions, and the overall band is the mean of the four skills rounded to the nearest half — with .25 rounding up to the next half band and .75 up to the next whole. So 6.125 becomes 6.0, 6.25 becomes 6.5, 6.75 becomes 7.0.

That arithmetic is less forgiving than it looks. Listening 7.5, Reading 7.0, Writing 5.5, Speaking 6.5 averages 6.625, which reports as 6.5 overall. Looks like a pass for a 6.5 requirement. It isn't, if the university also asks for a minimum of 6.0 in each component — and most of them do. One weak skill can sink an application that the overall band appears to satisfy.

There is no pass mark in IELTS. The institution sets the requirement, and the requirement usually has two halves: an overall band and a per-section floor. Read your actual offer letter, not the summary on an agent's website.

Pass condition: your lower of the two sittings is half a band above the overall requirement, and no single section is sitting on the floor.

Check 4 — can you trust the number at all?

A band you can act on and a band that's pure decoration look identical on the screen.

Listening and Reading are objective. Forty questions, a fixed answer key, a raw score out of 40 converted to a band using published tables. There's no judgement in it. So for Listening and Reading, a free mock is exactly as accurate as an expensive one, as long as it uses real Cambridge papers and the correct key. What money buys you elsewhere is the interface and the Writing feedback, not the arithmetic.

One caution on those tables: the Academic and General Training conversions are different, and General Training needs more correct answers for the same band. Any site that prints a single conversion chart without saying which module it covers is showing you the wrong number for one of the two modules.

Writing doesn't work like that. It's marked on four criteria, each carrying equal weight: Task Achievement (Task 1) or Task Response (Task 2), Coherence and Cohesion, Lexical Resource, and Grammatical Range and Accuracy. A Writing band is a human judgement about whether your position was clear throughout, whether your paragraphs each had a central topic, whether the less common vocabulary you reached for was used accurately. An algorithm can count subordinate clauses and average your sentence length. It cannot read your essay's argument. If a platform gives you a Writing band with no line-by-line reasoning against the four criteria, you have been given a number that means nothing, and acting on it is worse than having no number.

Apply those four criteria to real scripts and the same fault keeps surfacing. Most band-6 essays from strong candidates lose the mark on Coherence and Cohesion, not on grammar. The band 6 descriptor says the writer "may not always use referencing clearly or appropriately" and "uses paragraphing, but not always logically". Band 7 asks for a "clear central topic within each paragraph" and "clear progression throughout". That's structure, not tenses. The instinct at 6.5 is to buy another grammar book; the actual repair is usually one sentence per paragraph stating what that paragraph is for, and cutting the second body paragraph that repeats the first in new clothes.

The same descriptors demolish the advice you've been given about vocabulary. Band 6 Lexical Resource reads: "attempts to use less common vocabulary but with some inaccuracy". That sentence is a description of what "use advanced words" produces. Plethora, myriad, detrimental deployed slightly wrong pins you at 6 more reliably than plain accurate words would.

Our own limit, since we're claiming a human reads your script: humans disagree. Two competent markers can land half a band apart on the same essay. That's not a defect we can engineer away — it's why Check 3 exists and why the margin is half a band rather than zero.

Check 5 — Speaking, which a screen can only partly test

Say it straight: you can sit Speaking on our Cambridge papers, and you should know exactly what that is. You record your answers to all three parts, and an instructor listens and marks them on the four criteria, with a line on each. It is a person's judgement, not a microphone that returns a band it didn't earn, and it tells you where your Speaking stands.

What no on-screen mock can do, ours included, is put an examiner in the room. Speaking is a live interview — 11 to 14 minutes, three parts, with Part 2 giving you one minute to prepare and one to two minutes to talk from a cue card. The examiner follows up on what you actually said in Part 3. That interaction is the test. It's scored on Fluency and Coherence, Lexical Resource, Grammatical Range and Accuracy, and Pronunciation, and that fourth criterion doesn't exist anywhere else in the exam.

Do not let a strong Listening band stand in for it. Understanding an Australian accent on a recording and producing intelligible connected speech under pressure are different skills, and they come apart badly for a lot of Bangla speakers. The errors that survive into speech are specific and predictable: dropped articles ("I went to university near my house" becoming "I went to university near house"), advices and informations treated as countable, present perfect used with a finished time ("I have completed my HSC in 2021"), and on the pronunciation criterion, /v/ collapsing towards /w/ and /s/ towards /sh/. None of those show up in your Reading score. All of them are audible in ninety seconds.

Test it with a person. Sit someone opposite you who is willing to be blunt, hand them a Part 2 cue card, give yourself exactly one minute with a pencil and then talk until they stop you. Record it. The recording is the part that hurts and the part that works — you'll hear the four-second gaps you didn't notice, and the word you say eleven times.

When the honest answer is "not yet"

If your two sittings are a full band apart, or the lower one sits below your requirement, or your Writing band came from something that couldn't explain itself, the answer is wait a month.

A month is enough to move half a band if it's spent on diagnosis rather than volume. Build an error log from the two papers you already sat, categorised by question type rather than by skill — "Reading is weak" is useless; "I lose four marks a paper on Matching Headings and two on word-limit violations in sentence completion" tells you what to do on Tuesday. Spend the first fortnight on the two question types that cost you the most marks, timed at the section level, not the paper level. Rewrite one Task 2 you've already had marked, addressing the four criteria one at a time, and get it read again. Speak to a human being three times a week for fifteen minutes.

If you want that month laid out day by day rather than improvised, the 30-day plan is the same approach written as a schedule.

Then sit a third full paper at the end of the month. That's your new pair with the second one: same stability test, same margin rule.

When you book anyway

Sometimes the deadline decides. Work backwards: the date your university needs the score, minus 3-5 days for computer-delivered results (or about 13 for paper), minus the gap before a retake, minus the results wait on that retake. If the arithmetic only fits one sitting, book now, sit it as a real attempt, and treat it as the calibration you'd otherwise have paid a month for.

Before you book, ask your centre directly what retake options they run. Availability differs by country and changes, and I'm not going to tell you what's on offer in Dhaka this month when I don't know. If it is one skill holding the overall band down, One Skill Retake is worth pricing against a full resit before you decide.

The rubric, with its pass mark

You're ready when all five are true:

1. You've completed at least one full Listening, Reading and Writing block, timed, in exam order, on the format you'll sit.
2. You've done it twice, roughly a week apart, and the two bands are within half a band.
3. The lower band is half a band above your requirement, with no section resting on the per-component floor.
4. Your Listening and Reading bands came from a real answer key, and your Writing band came with reasoning against all four criteria.
5. You've been assessed speaking out loud by a person, not by proxy.

All five can start on a paper you sit properly — the real screen, real Cambridge material from books 1 to 21, Academic or General Training, with Listening and Reading marked the moment you submit and Writing and Speaking marked by an instructor. Every one of them is free, which is the point: a check you have to pay for is a check you'll put off until you already think you'll pass it — and by then it isn't telling you anything. What a mock costs at a centre in Bangladesh is the comparison that makes that argument concrete.