IELTS speaking vocabulary: a rare word can cost you two criteria and earn you none
Aman Batth · 10 min read · · Updated
Every vocabulary page hands you a list of advanced words to impress the examiner. From the marking side that is backwards — lexical resource is one of four equally weighted criteria, and reaching for a word you do not own damages at least two of them at once.
Search for IELTS speaking vocabulary and you will be handed a list. Two hundred phrases, sorted by part, with a better word offered for every ordinary one: good becomes rewarding, tired becomes exhausted, very important becomes paramount. The instruction is to learn them and deploy them.
The lists are well made. The advice around them is wrong, and it is wrong in a specific, measurable way.
We build the engine that scores speaking and writing answers on Hilingo, so this is written from the marking side. A word you do not fully own does not cost you one criterion when it goes wrong. It costs you two, and sometimes three. The official criteria themselves are published by the test owners — read them at ielts.org rather than taking any blog's word for the wording.
First, the arithmetic nobody puts next to the word list
IELTS speaking is marked on four criteria, weighted equally: fluency and coherence, lexical resource, grammatical range and accuracy, and pronunciation.
Vocabulary is one of the four. So it is a quarter of your speaking band, and speaking is a quarter of your overall band — which makes your entire vocabulary a sixteenth of the number on your test report.
That is worth holding on to, because it means the ceiling on what a word list can do for you is low. What it can cost you is not.
What reaching for a rare word actually does
Here is the same moment, scored across all four criteria. You are in Part 3, you want to say a country's economy is growing, and you have decided to use burgeoning because it was on a list.
| Criterion | Share of the speaking band | What the reached-for word does to it |
|---|---|---|
| Fluency and coherence | one quarter | You pause while you retrieve it. Retrieval hesitation is exactly what this criterion penalises. |
| Lexical resource | one quarter | Right, and you gain. Slightly off, and you lose — a near-miss scores below a plain word used correctly. |
| Grammatical range and accuracy | one quarter | Usually untouched — unless the word pulls a structure you do not control behind it. |
| Pronunciation | one quarter | A word learned from a page is a word you have never heard. You are guessing the stress live. |
Read the fluency row again, because it is the one the word lists never mention. The hesitation happens whether or not the word turns out to be correct. Even in the best case — you retrieve it, it fits, it lands — you have already paid for it in the criterion next door.
And read the pronunciation row, which is the quiet one. Vocabulary lists are read, not heard. Paramount, plethora, meticulous, inevitable all carry stress patterns you will not have practised if you met them in a table. Saying a word with the stress in the wrong place makes it less intelligible than the plain word you replaced, and intelligibility is what pronunciation is marked on.
So the honest version of the trade is this: an impressive word used well raises one criterion slightly. The same word used slightly wrong lowers one criterion, and the act of reaching for it lowers a second regardless.
What lexical resource actually rewards
Not rarity. Three things, and none of them are on a word list.
Precision. The criterion is about choosing the word that means the thing, not the word that sounds hardest. Exhausted is not a better tired — it is a different amount of tired, and using it for ordinary end-of-day tiredness is an imprecision, not an upgrade. A marker reads that as a candidate reaching, which is the opposite of what they are looking for.
Natural collocation. Words travel in company. Heavy rain, strong coffee, make a decision, take a risk. Collocation is where non-native fluency is most visible, and getting it right does more for lexical resource than any single rare word, because it is evidence you have the language rather than the vocabulary. This is the one part of the standard advice that is genuinely good, and it sits in every list under a heading nobody reads.
Flexibility — the ability to paraphrase. This is the big one, and almost nothing written about IELTS vocabulary covers it. The descriptors reward being able to say the thing when the exact word will not come. Not knowing more words. Getting around the missing one without stopping.
That last skill is the only one of the three that is trainable in weeks rather than years, and it is the one candidates spend no time on.
Paraphrase is a scored skill, not a fallback
Candidates treat "I didn't know the word" as a failure of preparation. In the marking it is a normal event with two possible outcomes.
Outcome one: you stop, say how to say, look up, stall, and eventually produce something. Fluency drops and lexical resource is not helped.
Outcome two: you go round it without breaking stride. I don't know the technical word for it, but it's the bit of the engine that... Fluency is intact, and you have just demonstrated the flexibility the criterion asks for.
Same gap in your vocabulary. Different band.
The drill: take any concrete noun — escalator, stapler, deadline — and describe it for fifteen seconds without using the word or any part of it. Then do it again faster. The skill you are building is not vocabulary; it is the reflex of not stopping when a word is missing. Five minutes a day is enough, and unlike a word list, it works on words you have never met, which is the only category that matters in the exam.
The register trap in Part 1
There is a second cost that lists create, and it comes from sorting phrases by part without saying what each part is for.
Part 1 is small talk. Do you like cooking? Do you use public transport? The correct register is conversational. A candidate who answers I find culinary pursuits immensely gratifying has not produced a band 8 sentence — they have produced a sentence that does not fit the question, and inappropriate register is a lexical resource problem, not a lexical resource achievement.
Part 3 is the abstract discussion, and it genuinely does take more formal, more precise language, because the ideas are more abstract. The vocabulary rises to meet the content. It does not get bolted on.
The test for any phrase you are thinking of preparing: would you say this to a colleague? If not, it will sound rehearsed in Part 1, and rehearsed is audible.
How to tell whether a word is actually yours
Everyone has two vocabularies: the one they recognise when reading, and the much smaller one they can produce under pressure. Word lists grow the first and leave the second where it was. The exam only marks the second.
A usable test, applied to any word you are considering:
- Produce it in three different sentences, out loud, without pausing. Not three sentences from the list — three of your own, about your own life.
- Say it without having to recall where you learned it. If the word arrives with its flashcard attached, it is not automatic yet, and under exam pressure it will arrive late or not at all.
- Know one thing it does not mean. Near-synonyms are where near-misses come from. If you cannot say what separates exhausted from tired, you do not yet control exhausted.
Anything that fails all three is a word you are gambling on. It might still be worth learning — over months, in context, by hearing it. It is not worth deploying next Saturday.
Idioms: the worst risk-to-reward on the list
Idioms deserve their own paragraph because they concentrate every problem above.
They are fixed, so a small deviation is not a variation but an error — once in a purple moon is not slightly wrong, it is wrong. They are register-specific, so the wrong one in Part 3 reads as flippant. They are the phrases most obviously drawn from preparation material, so a marker has heard a piece of cake and cost an arm and a leg many times that week. And they carry no more credit than any other appropriate lexical choice.
The advice most pages give — use one or two, naturally — is the right advice, and it quietly concedes the point. If the number is one or two, it was never the lever.
What to do with the weeks you have
Reordered from what the lists suggest, and deliberately short:
- Stop adding words. Start converting them. Take twenty words you already recognise and move them into production with the three-part test above. Twenty owned words outscore two hundred recognised ones.
- Drill collocation, not vocabulary. When you meet a word, learn the two or three words that sit beside it. Deadline is not the unit; meet a deadline, miss a deadline, tight deadline is.
- Practise the fifteen-second paraphrase daily. It is the only lexical skill that pays off on words you have not seen.
- Record yourself on Part 3 questions and listen for the reaches. You are looking for the half-second before a chosen word. That pause is the cost, on tape, and you cannot hear it while speaking.
- Prepare ideas for common topics, not sentences. Rehearsed sentences change the rhythm of your speech when they arrive, which is why memorised answers are detectable.
Nothing on that list will make a transcript of your answer look more impressive. All of it makes the recording score better, and the recording is what gets marked.
Find out which criterion your vocabulary is actually costing you
The reason this is hard to self-correct is the one that runs through all of it: you cannot hear your own hesitations, and you cannot see which of the four criteria a word choice landed in. An overall speaking band of 6.5 tells you nothing about whether the problem was the words or the pauses before them — and those have opposite fixes.
Hilingo gives you one full scored mock free, with no card. Speaking is assessed across Parts 1, 2 and 3 and comes back with a band on each of the four criteria separately — fluency and coherence, lexical resource, grammatical range and accuracy, pronunciation — rather than as a single number. Writing is marked the same way, on all four band descriptors. That is what tells you whether your vocabulary is holding you back or whether reaching for it is.
- The IELTS mistakes you cannot see yourself — why one criterion is usually a band below the rest
- The last week before your IELTS — why late vocabulary work rarely pays
- IELTS revaluation: do the arithmetic first — speaking is one of the two sections worth challenging
- IELTS band score chart — band to CLB per skill
Frequently asked questions
Do I need advanced vocabulary for IELTS speaking?
Less than you have been told. Lexical resource is one of four equally weighted speaking criteria, so your whole vocabulary is a quarter of the speaking band. What the criterion rewards is precision, natural collocation and the ability to paraphrase — not rare words. A plain word used exactly right scores better than an advanced one used approximately.
Does using difficult words increase your IELTS speaking band?
Only if you use them accurately and without hesitating. A word you have to retrieve costs you on fluency and coherence whether or not it turns out to be correct, and if it is slightly off it lowers lexical resource as well. That is two criteria at risk for a small gain in one.
What is lexical resource in IELTS speaking?
It is the criterion covering your range of vocabulary and how appropriately you use it. In practice it assesses whether you choose words that precisely fit your meaning, whether your word combinations sound natural, and whether you can paraphrase when the exact word does not come to you. The published descriptors are on ielts.org.
How many idioms should I use in IELTS speaking?
At most one or two, and only ones you use in ordinary conversation already. Idioms are fixed phrases, so a small deviation is an outright error rather than a variation, and the common ones are instantly recognisable as preparation material. They carry no more credit than any other appropriate word choice, which makes their risk-to-reward the worst on any vocabulary list.
Is it bad to repeat the same word in IELTS speaking?
Repeating one word many times does limit the range a marker can see, but the fix is not a synonym you are unsure of. Substituting a word you do not fully control is more damaging than the repetition was. Build range in words you can already produce under pressure, and paraphrase the idea rather than swapping the word.
How can I improve my vocabulary for IELTS speaking in a month?
Convert rather than collect. Take words you already recognise and practise producing each in three of your own sentences out loud without pausing, learn the two or three words that naturally sit beside each one, and spend five minutes a day describing objects without naming them. A month is not long enough to make new words automatic, but it is long enough to make existing ones reliable.