How to practise IELTS Speaking alone (without a partner)

Most people who ask this question assume that practising alone is the compromise option — the thing you do because you cannot find a partner. It is not. The IELTS Speaking test is not a conversation between equals; it is eleven to fourteen minutes of one person asking and one person answering, to a script the examiner cannot deviate from. Almost everything that decides your band — how long you keep going, how precise your words are, whether your grammar holds up under time pressure — is something you can build on your own, with nothing but a phone and a timer.

What solo practice genuinely cannot do is tell you where you stand. That is the gap worth planning around, and it is the last section of this guide.

Why practising alone works better than most people expect

A speaking partner gives you two things: someone to talk to, and someone to react. Only the first matters much for IELTS, and you can get it from a recording device. The second — a partner’s opinion of your English — is usually worth less than people hope, because a partner at a similar level hears the same things you do and misses the same things you miss.

Meanwhile, practising alone gives you something a partner cannot: the ability to repeat. You can answer the same cue card four times in twenty minutes, changing one thing each time, and listen to all four. No partner will sit through that, and it is the single most effective drill there is.

Record yourself, then listen back twice

This is the foundation, and it is the step people skip. Open the voice recorder your phone already has, answer one question, stop.

Then listen back twice, with a different job each time.

The first pass is about content. Did you answer the question that was asked, or the question you wished had been asked? Did you give a reason, an example, a contrast — or did you list facts until you ran out? Most Band 6 answers are not short of vocabulary; they are short of ideas developed past the first sentence.

The second pass is about language, and here you count rather than judge. Count the times you said “um” or restarted a sentence. Count how many times the same adjective came back. Count the sentences that had only one clause. Counting gives you a number you can beat next week; “that sounded a bit weak” gives you nothing.

The first few recordings are unpleasant to listen to. Everyone’s are. That discomfort passes after roughly three sessions, and what is on the other side of it is the most honest feedback available to you without paying someone.

Use a cue card with a real timer

Part 2 is the part where practising alone comes closest to the real thing, because it genuinely is a monologue. You get the card, one minute to make notes, then one to two minutes of uninterrupted talking.

Do it with the actual clock. A minute of preparation is shorter than it sounds, and two minutes of talking is much longer — most people dry up at about seventy seconds the first time and are shocked by it. The cue card library has a timer built into every card for exactly this, so you are not juggling a stopwatch and a topic at once.

Then repeat the same card. Second attempt: same story, but force three complex sentences into it. Third attempt: same story again, but replace every vague word (“good”, “nice”, “a lot”) with a precise one. By the third run the story is automatic and all your attention is on language, which is the state you want to reach before test day. There is a fuller method for building the two minutes in our guide to answering any cue card.

Sit a full mock test end to end

Individual answers are not the whole skill. On the day, the difficulty is cumulative: Part 1 warms you up, Part 2 tires you out, and then Part 3 asks you to be abstract and analytical when you are least fresh. People who have only ever practised single answers routinely fall apart in Part 3 for that reason alone.

So run the whole thing, timed, without stopping — the mock test does this in the browser: Part 1 questions, a cue card with the one-minute prep, then the Part 3 discussion that follows from it, in order, with no pauses for you to collect yourself. Do it once a week rather than every day. The point is not to grind; it is to make the shape of the test boring, so that on the day the only new thing in the room is the examiner.

Work through Part 1 and Part 3 question banks

Between mocks, the useful unit is the topic, not the single question. Examiners move through a topic in a predictable arc: Part 1 asks about you and your habits in the present, Part 3 takes the same subject and turns it abstract — causes, comparisons, consequences, what might change.

Working through questions by topic trains both halves at once and surfaces the vocabulary gaps that only appear under a specific subject. You will discover you can talk fluently about food and go blank on transport, which is exactly the kind of thing you want to find out in your kitchen and not in the test room.

For Part 3 specifically, practise the shape of the answer rather than the content: a position, a reason, an example, then a concession. Say it out loud four times with four different topics and it becomes a reflex.

Score yourself, carefully

You can assess your own speaking, but only structurally, and only one criterion at a time. Trying to give yourself an overall band from a single listen is how people convince themselves they are a 7 for six months.

Take one recording and listen for grammar alone. Then listen again for vocabulary alone. Then fluency. Write a single sentence about each. The self-assessment turns this into twelve concrete questions about how you speak and produces a band range rather than a single number, which is the honest form for an estimate.

The range matters more than most people realise, because the real examiners produce one too — two examiners marking the same answer can land half a band apart, which is the subject of why your score is a range. Train against the strict end of your own range and a generous examiner on the day can only work in your favour.

Shadowing, and what it is actually for

Shadowing — playing a recording of a fluent speaker and speaking along with it a half-second behind — is the one solo technique that directly improves pronunciation and rhythm. It works because you are copying stress and intonation in real time rather than thinking about them.

Keep its scope honest, though. Shadowing does not build ideas, range or grammar; it builds the sound. Ten minutes of it inside a session is plenty, and it is best spent on sentence stress, because sentence stress is what makes English sound fluent even when the individual sounds are accented.

When an outside score is worth it

Here is the limit of doing this alone: you cannot hear your own errors if you do not know they are errors. A grammar mistake you have made for ten years sounds correct to you. That is the whole problem — it is fossilised, and no amount of listening back will flag it, because your ear has already accepted it.

This is the point at which an outside judgement earns its place. That can be a tutor, if you have access to one, or it can be an app that marks the answer against the four criteria and names the specific pattern that keeps recurring — which is what IELTS Mentor AI was built to do. Either way, the useful cadence is periodic, not constant: get a scored answer, find out which two things are actually costing you marks, then go back to practising alone for a fortnight with those two things as your only focus.

That rhythm — long stretches of solo repetition, punctuated by an outside check — is how almost everyone who improves quickly actually works. The partner was never the missing piece.