IELTS Speaking Practice: Prepare and Score Higher

The IELTS speaking test format

The IELTS speaking test is a short face to face conversation with a trained examiner, lasting about eleven to fourteen minutes, split into three parts. It is recorded, and it is the same test whether you sit the Academic or the General Training version. You talk, the examiner listens and gently steers, and behind the friendly chat they are scoring your English against a fixed set of criteria the whole time.

It helps to picture the shape of it before you worry about any single answer. Part 1 is a warm up interview about you and everyday topics, made of short back and forth questions. Part 2 is the long turn, where you get a card with a topic, one minute to make notes, and then you speak alone for up to two minutes. Part 3 is a two way discussion that grows out of the Part 2 topic but pushes into more general, more abstract territory, with the examiner asking follow up questions and challenging your ideas a little.

One thing surprises a lot of test takers: the examiner is not trying to trip you up. They follow a script for the questions but they are on your side in the sense that they want to give you enough room to show what you can do. The test is short, so every second of talking is a chance to demonstrate range. Silence, one word answers, and long panicked pauses all waste that chance. The single most useful mental shift is to treat the whole thing as a conversation you are trying to keep going, not an exam you are trying to survive.

Because the format never changes, you can practice the exact structure until it feels routine. That familiarity is worth real marks on the day, because nerves eat fluency, and knowing precisely what is coming in each part takes a large chunk of the nerves away. If you have ever frozen when someone asked you a simple question in English, the fix is not more grammar, it is more reps of actually answering out loud, which is the thread running through this entire guide and through our wider piece on English speaking practice.

Part 1: the interview

Part 1 lasts roughly four to five minutes and is a set of short questions about you and familiar topics such as your home, your work or studies, your hobbies, food, weather, or daily habits. The examiner works through a few topic areas, asking a handful of questions in each. Answers here are meant to be fairly short, a sentence or two or three, not a speech, but short does not mean one word.

This part is easy to underrate. People assume that because the questions are simple, the answers do not matter much. They matter a lot, because Part 1 is where the examiner forms a first read of your fluency and your ease with the language, and it sets the tone for everything after. The trap is answering like you would tick a box. Asked "Do you like cooking?" the low scoring answer is "Yes." The band raising answer adds a reason and a small detail: "Yes, I really enjoy it, especially on weekends when I have time to try a new recipe instead of just throwing together something quick after work."

See what that longer answer does. It gives an opinion, a reason, a time reference, and a natural contrast, all in one breath, and it sounds like a person talking rather than a form being filled in. That is the model for Part 1: answer the actual question, then extend it by one notch with a because, an example, a comparison, or a feeling. You are not padding, you are giving the examiner material to score.

At the same time, do not overshoot. Part 1 is not the place for a two minute monologue, and rambling on until the examiner has to cut you off works against you. The rhythm you want is a genuine answer plus a bit of colour, then stop and let the next question come. Think of it as a comfortable chat with someone you have just met who is politely curious about your life. Warm, natural, a little more than the bare minimum, and then a clean handover back.

Part 2: the long turn

Part 2 is the long turn, and it is the part most people fear. The examiner hands you a card with a topic and a few prompts, gives you paper and a pencil, and you have exactly one minute to prepare. Then you speak on your own for one to two minutes, and the examiner will usually let you run to the full two minutes before stopping you and asking one or two quick rounding off questions.

The card almost always follows the same pattern. It names something you should describe, a place, a person, an object, an event, a plan, then lists a few points to cover, usually along the lines of what it is, when or where, why it matters to you, and how you feel about it. Because the structure is so predictable, your one minute of prep should not be spent panicking about ideas. It should be spent jotting a few keywords next to each prompt so you have a path to follow when you start talking and never dry up halfway.

The real skill in Part 2 is talking continuously for two minutes without stalling, repeating yourself, or trailing off into silence. Two minutes feels long when you are alone and the clock is running. The way to fill it well is to treat each prompt on the card as a mini section and to keep asking yourself the extending questions as you speak: why, how, what happened next, how did that make you feel, what would you compare it to. A concrete example beats a vague statement every time. Instead of "It was a nice place," you say what you could see, hear, and smell, and who you were with, and why it stuck in your memory.

Notes are your safety net, so use the prep minute well. You are not writing sentences, you are writing triggers, three or four words per prompt that will each carry thirty seconds of speech. A common way to lose marks here is to prepare too little and then stop after forty seconds because you have run out of things to say. The other common way is to prepare a rigid script in your head and then sound like you are reciting. Aim for the middle: enough of a plan to keep moving, loose enough to sound like you are thinking out loud, which you should be.

Part 3: the discussion

Part 3 is a four to five minute two way discussion, and it takes the topic from your Part 2 card and lifts it into broader, more abstract questions. If Part 2 asked you to describe a skill you learned, Part 3 might ask why some skills are harder to learn as an adult, whether schools should teach practical skills, or how technology is changing the way people learn. The focus shifts from your own life to opinions, comparisons, causes, and predictions about people and society in general.

This is where the higher bands are won or lost, because the questions demand real thinking in English, not just describing familiar things. The examiner is looking for whether you can develop an argument, weigh two sides, give examples, speculate about the future, and handle a follow up that pushes back on what you just said. You do not need expert knowledge of any topic. You need to be able to reason out loud clearly, and to structure a spoken answer so a listener can follow your logic.

A reliable way to build a strong Part 3 answer is to state your position, give a reason, back it with an example, and then acknowledge the other side or a limit to your view. That four step move, claim, reason, example, balance, works for almost any question and stops your answers from being a single flat sentence. It also naturally showcases the linking language, the range of tenses, and the opinion vocabulary that the band criteria reward, without you having to force any of it.

Expect to be challenged, and welcome it. If the examiner says "But some people would disagree, what would you say to them?" that is not a sign you got it wrong, it is an invitation to show you can handle a harder turn. Concede the reasonable part of the other view, then explain why you still hold yours. Speculating is fine and often expected: "It is hard to say for certain, but I would imagine that over the next decade..." is exactly the kind of hedged, thoughtful language a strong candidate produces. Part 3 rewards the person who can think and talk at the same time, which, again, only comes from practice.

The four band criteria explained

Your speaking score comes from four criteria, each weighted equally at a quarter of the total, then averaged into your overall speaking band from 0 to 9. Because they carry equal weight, one weak area pulls down an otherwise strong performance, so it pays to know exactly what each one measures. Here they are, described the way the public band descriptors describe them, in plain language.

Fluency and coherence

This measures how smoothly and how logically you speak. Fluency is about flow: can you talk at a natural pace, at length, without constant pauses, and without leaning on fillers like um, er, and like to buy time. Coherence is about sense: do your ideas connect, do you use linking words well, can a listener follow your train of thought from one sentence to the next. A higher band here means you speak at length, hesitate mainly to find ideas rather than words, and connect your points clearly.

Two things people misunderstand. First, a pause is not automatically bad. Pausing briefly to think about what you want to say is normal and human, and the descriptors treat it very differently from pausing because you cannot find the language. Second, speaking fast is not the same as being fluent. Racing through your answer to seem confident often hurts you, because it wrecks your clarity and your pronunciation. Steady, connected, and clear beats fast and messy every time.

Lexical resource

This is your vocabulary: the range of words you can use, how precisely you use them, and whether you can paraphrase when you do not know an exact word. A strong performance uses a good spread of vocabulary, including some less common words and natural expressions, mostly in the right context, and can talk around a gap smoothly instead of freezing when the perfect word will not come. You do not need rare, showy words. You need the right word for the meaning, used naturally.

The most useful skill inside this criterion is paraphrasing. When you cannot remember the exact term for something, a lower scoring candidate stops dead or switches to their own language. A higher scoring one describes it: "the thing you use to open a bottle" instead of the missing word, and keeps going. Being able to reroute around a gap protects your fluency and shows flexibility, which is precisely what the descriptors reward. Cramming a list of fancy words the night before does the opposite, because you tend to drop them in awkwardly and they stick out.

Grammatical range and accuracy

This looks at two things together: the range of grammar you use, and how accurate it is. Range means using a mix of sentence types, simple, compound, and complex, and handling different tenses and structures rather than sticking to short, safe, identical sentences. Accuracy means how often those sentences are error free. The strongest performances use a variety of complex structures with only occasional slips that do not block understanding.

There is a real tension here that you have to manage. If you only ever speak in short simple sentences to stay safe, your accuracy might be high but your range is poor, and range is half the score. If you attempt very complex sentences everywhere, you show range but pile up errors. The sweet spot is to mix confidently: use complex structures where you can control them, keep some simple sentences for clarity, and do not tie yourself in knots trying to be fancy. Controlled variety, not error free simplicity and not ambitious chaos.

Pronunciation

This measures how easy you are to understand. It covers individual sounds, word stress, sentence stress, rhythm, and intonation. The key idea, and this is where many candidates worry needlessly, is that the goal is clear communication, not a British or American accent. You can keep your accent completely and still score well, as long as your speech is easy to follow and you use stress and intonation in a way that helps meaning come through.

What actually costs marks is when pronunciation makes you hard to understand: sounds that blur into each other, stress placed on the wrong syllable so a word sounds unfamiliar, or flat, robotic intonation that hides where the important points are. The fixes are physical and specific, and they respond well to practice. If pronunciation is your weak criterion, our guides on how to improve pronunciation and accent reduction go deep on the exact habits that make English speech clearer without erasing who you are.

Notice the common cure across all four rows: speak out loud, often, and review yourself. No amount of silent reading moves any of these four numbers, because every one of them is about production, and production only improves by producing.

Sample question types

You cannot know the exact questions in advance, but the question types are stable and you can prepare for every one of them. Knowing the shapes lets you build flexible answers rather than memorizing scripts, which, as covered later, backfires. Here is what each part tends to throw at you.

Part 1 question types

Short, personal, and about familiar topics. They cluster around a few themes and reuse the same question stems:

• Simple preference and habit: "Do you enjoy...?", "How often do you...?", "What kind of... do you like?" Answer with an opinion plus a reason.
• Description of your life: "Tell me about your hometown," "What do you do for work or study?", "Describe your typical morning." Give a couple of specific details.
• Past and future: "Did you enjoy... as a child?", "Would you like to... in the future?" These let you show different tenses naturally.
• Comparison: "Do you prefer... or...?" A chance to weigh two things briefly and explain your choice.

Part 2 cue card types

The long turn card asks you to describe something, and the something is almost always a person, place, object, event, experience, or plan. Typical prompts read like these:

• Describe a person who has influenced you.
• Describe a place you like to visit.
• Describe a skill you would like to learn.
• Describe an occasion when you helped someone.
• Describe a piece of technology you find useful.

Under each you will find the same style of sub prompts, telling you to say what it is, when or where, some detail, and how you feel about it. Because the categories repeat, you can build a small stock of personal stories, a memorable trip, a person you admire, a hobby, a challenge you overcame, that can be bent to fit many cards. That is preparation, not memorization, because you are still describing the specific card in front of you.

Part 3 question types

Abstract and opinion based, linked to the Part 2 theme. The stems tell you what kind of thinking is wanted:

• Opinion: "Do you think... is a good idea?" State a position and defend it.
• Comparison and contrast: "What are the differences between... and...?" Weigh both sides.
• Cause and effect: "Why do you think... has become more common?" Give reasons.
• Speculation about the future: "How might... change in the next twenty years?" Use hedged, tentative language.
• Problem and solution: "What can be done about...?" Suggest and justify.

For every one of these, the claim, reason, example, balance move gives you a ready structure. You are never expected to be right in some absolute sense. You are expected to hold a view and support it in clear, connected English.

How to practice each part

General advice is comforting and useless. What moves your band is targeted practice on each part under the real conditions, followed by honest review. Here is how to drill each one so your practice actually mirrors the test.

Practicing Part 1

Gather common Part 1 topics, home, work, study, hobbies, food, travel, weather, technology, and answer the questions out loud, not in your head and not in writing. Record yourself. Then play it back and check one thing: did you answer plus extend, or did you give bare one line replies. Re do any answer that came out too short until adding a reason or a detail feels automatic. The aim is that a natural, slightly extended answer becomes your default, so you never have to think about it on the day.

Practicing Part 2

This is the part that most rewards realistic drilling, because the timing is what makes it hard. Always practice with a strict clock: exactly one minute of prep with pen and paper, then speak for the full two minutes while recording. Do not stop early even if you feel you have run out, push to keep talking, because that is the exact skill the part tests. Afterward, listen back and mark where you stalled, repeated yourself, or ran dry, then do the same card again and beat your last attempt. A handful of cue cards done this way, with real timing and real review, is worth more than reading fifty sample answers.

Practicing Part 3

Take abstract questions and answer them out loud using the claim, reason, example, balance structure until it is second nature. The hard part of Part 3 is thinking and speaking at once, so the practice has to involve real time pressure, not the luxury of drafting a perfect answer on paper. This is where a responsive partner helps most, because a good Part 3 question leads to a follow up, and you need reps at handling the unexpected push back rather than delivering a rehearsed piece. If you have no partner handy, an AI one that actually replies and challenges you fills the gap, which is a large part of why speaking apps exist.

Common mistakes that cost band points

Most lost marks are not from lacking English, they are from habits that undercut the English you have. These come up again and again, and every one of them is fixable with awareness and practice.

1. Answers that are too short. Especially in Part 1, one word and one line answers give the examiner nothing to score and make you sound reluctant. Always answer, then extend by one notch. This is the most common and most fixable mistake of all.
2. Memorizing and reciting. Examiners are trained to catch memorized speeches, and a recited answer sounds unnatural and often does not fit the exact question, which hurts your coherence. Learn flexible language, not fixed scripts.
3. Overusing fillers. A constant stream of um, er, like, you know, and basically drags your fluency score down. A short, silent pause to think reads far better than a noise. Practice being comfortable with brief silence.
4. Speaking too fast. Rushing to sound confident blurs your pronunciation and coherence. A steady, clear pace lets all four criteria show. Slow down, especially when you start.
5. Going off topic. Answering a slightly different question than the one asked, often because you steered toward a memorized answer, costs coherence marks. Listen to the actual question and answer that.
6. Playing it too safe with grammar. Only using short simple sentences keeps you accurate but tanks your range, which is half that criterion. Mix in complex structures you can control.
7. Trying to be too fancy. The opposite error: forcing rare words and tangled sentences you cannot handle, which piles up errors and hesitation. Aim for natural and correct over impressive and shaky.
8. Freezing on an unknown word. Stopping dead when you cannot find a word wrecks your flow. Paraphrase around it and keep moving. Talking around a gap is a scored skill, not a failure.
9. Ignoring pronunciation. Treating pronunciation as unfixable and hoping the other three carry you. It is a quarter of your score, and stress and intonation respond quickly to focused practice.
10. Not practicing out loud. Reading tips and watching videos while never actually speaking. Speaking is a physical skill, and you cannot build it silently. Reps are everything.

Look at the pattern. Almost every one of these is a habit, not a knowledge gap, which is good news, because habits change with deliberate, repeated practice. Fix the short answers, drop the memorized scripts, calm your pace, and get your voice moving every day, and you claw back marks that had nothing to do with how much English you actually know.

Timing and pacing on the day

Knowing the clock removes a surprising amount of stress, so fix the rough timings in your head before you walk in. The whole test runs about eleven to fourteen minutes. Part 1 takes four to five minutes of short questions. Part 2 is a strict structure: one minute to prepare, then one to two minutes to speak, with the examiner usually stopping you at two. Part 3 fills the remaining four to five minutes of discussion. You do not control the timing, the examiner does, but knowing it lets you pace yourself.

The pacing mistakes are predictable. In Part 1, some people give speeches and get politely cut off, which is fine but a sign to trim. In Part 2, the danger is the opposite, running dry before two minutes, so your job is to keep expanding until you are stopped. In Part 3, rushing your answers to seem sharp actually reduces how much good language you produce, because a rushed answer is usually a thin one. A calm, slightly slower pace gives you room to structure your points and lets your pronunciation stay clean.

A few practical pacing habits pay off on the day. Take the first couple of seconds of any answer to breathe and think rather than blurting, because a two second gather at the start prevents a scramble halfway through. In Part 2, glance at your notes to move between prompts so you never stall between sections. And throughout, remember that being stopped by the examiner is neutral, not negative, it simply means the part is over, so never read a cut off as a sign you did badly.

Because the timings are fixed, you can rehearse against them exactly. Set a timer for your Part 2 practice every single time, and get used to the feel of one minute of prep and two minutes of talk until your internal clock matches the real one. On test day, the timing then feels familiar instead of alarming, and familiarity is the cheapest fluency boost available, since it frees up the mental space that nerves would otherwise eat.

A daily practice routine

Consistency beats intensity for speaking, because fluency is a physical habit built from many small reps, not a fact you can cram. A focused twenty to thirty minutes most days for a few weeks does far more than a single marathon session the weekend before. Here is a simple daily routine that touches all three parts and all four criteria, and flexes shorter on busy days.

The loop this creates is what makes it work. You warm up your voice, drill the hardest part under real timing, practice thinking out loud, then review yourself and feed tomorrow a small target. Over a few weeks the fillers thin out, the short answers grow reasons, the two minute talk stops feeling long, and reasoning in English stops feeling like translation. None of that comes from a single heroic study day. It comes from showing up for a short session most days, which is the whole point.

Adjust the balance to your weak spot. If Part 2 is your terror, give it more of the time and less to the warm up. If pronunciation is your low criterion, add a couple of minutes of reading a passage aloud with exaggerated stress and intonation, and compare your recording to a native sample. If you are already comfortable and just polishing, lean into Part 3, since that is where the higher bands are decided. And whatever you do, protect the daily-ness over the length, because a short session every day beats a long one once a week for a skill that lives in habit. For the broader habits behind speaking freely, our guide on English speaking practice pairs well with this routine.

Why solo practice with feedback works

Here is the honest problem with preparing for a speaking test: speaking needs a listener, and a patient, available, judgment free listener is hard to find. You can read every tip, watch every sample video, and memorize every band descriptor, and still walk in unable to speak fluently, because none of that is speaking. The test measures production, and production only improves by producing, out loud, again and again.

Most people fall into one of two traps. They practice silently, running answers in their head, which feels like work but trains almost nothing, because thinking an answer is far easier than saying it under time pressure. Or they wait for a study partner or a tutor slot, which means they practice rarely, in short bursts, on someone else's schedule, when the whole method depends on frequent daily reps. Both traps lead to the same place: a candidate who knows a lot about the test and has barely rehearsed the actual skill it grades.

Feedback is the other missing half. You cannot fully hear your own speech while you produce it, because your attention is busy forming the next sentence. Your own ear forgives your own fillers, your own short answers, your own flat intonation, precisely because they are yours and they sound normal to you. Recording yourself helps a great deal, and you should do it, but recordings are passive, they show you the problem without telling you what to do about it or giving you a fresh prompt to try again on.

This is exactly the gap an AI speaking partner closes, and it is why solo practice with feedback has become such a practical way to prepare. It is available whenever you are, so you actually get the daily reps the skill needs. It never gets bored of your mistakes, so nervous speakers finally produce out loud instead of hiding in silent study. It responds in real time, which trains the Part 3 skill of handling an unexpected turn, and it can point at what to fix in your fluency and pronunciation rather than leaving you to guess. That combination, real spoken reps plus immediate feedback, is what actually moves the four numbers, and it is available on your own timetable rather than someone else's.

How SPAIKING fits your IELTS prep

SPAIKING is an AI app you talk to, out loud, in real spoken conversation, and that makes it a natural fit for the one thing IELTS speaking prep most often lacks: reps. You cannot practice a speaking test by reading, and you rarely have a willing partner at the exact moment you have twenty free minutes. SPAIKING is there whenever you are, which means the daily routine above stops being a nice idea and becomes something you can actually do every day.

Here is how it slots into the parts of the test. For Part 1, you hold a natural back and forth on familiar topics and practice answering plus extending until it is automatic. For Part 2, you rehearse the long turn out loud, working on filling the full two minutes without stalling, since the fix for running dry is simply having done it many times. For Part 3, because the AI actually responds and can push back, you build the habit of reasoning in spoken English and handling a follow up you did not see coming, which is the exact pressure a static list of sample answers can never give you.

The reason it helps is not magic, it is the two ingredients from the last section put together:

• It is always available. No scheduling a partner, no waiting for a tutor slot. Open it and speak, which is how you get the daily reps that fluency demands.
• It is judgment free. You can fumble the same answer ten times and the AI just keeps the conversation going, which is what finally gets nervous test takers to produce out loud instead of rehearsing in their heads.
• It gives feedback in real time. Your own ear forgives your fillers and pronunciation slips; immediate correction points you at what to fix while it is still fresh.
• It responds and challenges. Because it talks back, you practice the unscripted, two way pressure of Part 3, not just the delivery of a memorized piece.

None of this replaces knowing the format or understanding the criteria, and this whole guide exists to give you that. But once you know what the test wants, the work that remains is speaking, a lot, with feedback, on a schedule you can keep. That is the half people skip, and it is the half that decides your band. You can see plans on our pricing page, and you can always start from the homepage. Prepare with the test's shape clear in your head, then put in the spoken reps, and you walk into the room having already done, many times over, the exact thing the examiner is about to ask of you.