PTE vocabulary: the score guide asks for the right word, not a rare one
Aman Batth · 23 min read · · Updated
Vocabulary is a named trait on four question types — four of PTE Academic's twenty-two — and on three of the four the top band says appropriate, not advanced. That is the opposite of what a word list trains — which is why a rare word used slightly wrong costs you more than a plain word used correctly.
Search for PTE vocabulary and you get lists. Twenty-five academic words with synonyms and antonyms. Eighteen categories of linking phrases. A pronunciation guide. All of it perfectly accurate, and almost none of it connected to how a single mark is awarded.
Here is the fact those pages skip. In Pearson's Test Taker Score Guides, Vocabulary is a named trait on four question types — four of PTE Academic's twenty-two — and on three of the four, the top band asks for the right word, not an impressive one. The exact phrase "appropriate choice of words" appears in two of them; a third, Answer Short Question, says "appropriate word choice in response."
That is not a small distinction. It means a rare word used slightly wrong damages the trait, while a plain word used correctly scores the maximum. A list of advanced synonyms trains you to do the first thing.
We build the engine that marks speaking and writing answers on Hilingo, so this is written from the marking side. Everything below is checkable in the PTE Academic Test Taker Score Guide and the PTE Core Score Guide, and I have given the page numbers so you can.
Where the word "Vocabulary" actually appears in the scoring
Read the "Traits scored" column of the scoring criteria tables and it appears four times per test. Nowhere else.
PTE Academic
| Question type | No. | Vocabulary trait | Out of | Guide page |
|---|---|---|---|---|
| Answer Short Question | 5–6 | Vocabulary | 0–1 | 24 |
| Summarize Written Text | 2 | Vocabulary | 0–2 | 33 |
| Write Essay | 1 | Vocabulary Range | 0–2 | 39 |
| Summarize Spoken Text | 1 | Vocabulary | 0–2 | 43 |
PTE Core
| Question type | No. | Vocabulary trait | Out of | Guide page |
|---|---|---|---|---|
| Answer Short Question | 5–6 | Vocabulary | 0–1 | 30 |
| Summarize Written Text | 1–2 | Vocabulary | 0–2 | 31 |
| Write Email | 2–3 | Vocabulary | 0–2 | 33 |
| Summarize Spoken Text | 1–2 | Vocabulary | 0–2 | 37 |
Now the column for everything else.
Every reading question type on both tests reads "Not applicable" under Traits scored (Academic guide page 40, Core page 35). Fill in the Blanks, both varieties. Multiple Choice. Reorder Paragraph. No traits at all — they are scored one point per correct blank, response or adjacent pair, except Multiple Choice, Multiple Answers, which also deducts one point per incorrect selection, with the question floored at zero.
Every objective listening question type reads "Not applicable" too (Academic page 44, Core page 38). Including Write from Dictation, which is scored one point per word spelled correctly and has no vocabulary trait on it whatsoever.
And every speaking task on PTE Academic is scored on Content, Pronunciation and Oral Fluency. Read Aloud, Repeat Sentence, Describe Image, Retell Lecture, Respond to a Situation, Summarize Group Discussion. Not one of them carries a Vocabulary trait. PTE Core has neither Retell Lecture nor Summarize Group Discussion, and its Respond to a Situation is scored on Appropriacy, Pronunciation and Fluency — Appropriacy standing in place of Content, and about register rather than vocabulary.
If you have been drilling advanced synonyms for Describe Image, that is the first thing worth knowing. There is no vocabulary mark on that task to win.
(This is the same map as PTE marks distribution, read down a different column. If you have not worked out which of your four skill scores is actually capping you, start there instead — it decides whether any of this is where your marks are.)
Three of the four top bands say "appropriate". None says "advanced"
This is the part that changes what you should practise, so it is worth reading the descriptors as written rather than paraphrased.
Summarize Written Text, Vocabulary, Academic page 33 and Core page 31:
2 Has appropriate choice of words 1 Contains lexical errors but with no hindrances to communication 0 Has defective word choice which could hinder communication
Summarize Spoken Text, Vocabulary, Academic page 43 and Core page 37:
2 Appropriate choice of words 1 Some lexical errors but with no hindrance to communication 0 Defective word choice which could hinder communication
Answer Short Question, Vocabulary, Academic page 24 and Core page 30:
1 Appropriate word choice in response 0 Inappropriate word choice in response
Three of the four. The scale is not basic to advanced. It is correct, then one error, then errors that hinder meaning. There is no band above "appropriate". Producing a perfectly ordinary word that fits the sentence earns the maximum on the trait — two points on the two summary tasks, one on Answer Short Question — and there is nothing further to earn.
The fourth one is the exception people generalise from, and it is worth reading carefully too.
Write Essay, Vocabulary Range, Academic page 39:
2 Good command of a broad lexical repertoire, idiomatic expressions and colloquialisms 1 Shows a good range of vocabulary for matters connected to general academic topics. Lexical shortcomings lead to circumlocution or some imprecision 0 Contains mainly basic vocabulary insufficient to deal with the topic at the required level
This one does reward breadth. It is also the only one that does, it appears on a task that occurs once in a PTE Academic test, and it is worth two of the twenty-six trait points on that essay. The essay is scored on Content 0–6, Form 0–2, Development Structure and Coherence 0–6, Grammar 0–2, General Linguistic Range 0–6, Vocabulary Range 0–2 and Spelling 0–2.
And look at what sits beside it. General Linguistic Range is worth 6, three times as much, and its top band reads:
6 A variety of expressions and vocabulary are used appropriately to formulate ideas with ease and precision throughout the response. No signs of limitations restricting what can be communicated. Errors in language use, if present, are rare and minor, and meaning is completely clear.
"Appropriately." "Precision." "Errors... rare and minor." The six-point trait that governs range makes range conditional on accuracy. You cannot buy band 6 with vocabulary you do not control, because errors in language use are written into the same descriptor.
Pearson does not publish how trait points aggregate into the 10–90 skill scores, so nobody — us included — can tell you what two Vocabulary Range points are worth in score terms. What the rubric does tell you is the ordering, and the ordering is unambiguous.
What this does to a rare word you half-know
Put the two halves together and the arithmetic falls out.
A plain, correct word on Summarize Written Text: Vocabulary 2. Nothing lost anywhere.
A rare word used slightly wrong — right register, wrong collocation, or the right idea in a word that does not take the preposition you gave it: that is a lexical error. Best case it drops Vocabulary to 1. If the sentence now hinders the reader, Vocabulary is 0. And because word choice errors are frequently also grammar errors (wrong complement pattern, wrong countability, wrong part of speech), the same word can land on the Grammar trait beside it, which is also worth 0–2.
So the exchange rate is this. A correct plain word scores 2 and risks nothing. An incorrect rare word scores at most 1, often 0, and can take a second trait with it. There is no upside band to compensate, because on three of the four vocabulary traits, the top band is already what "ordinary and correct" earns.
Against that, on the one task where breadth pays — Write Essay — the gain is capped at two points and is conditioned on the six-point range trait that penalises the same error.
That is the opposite of what memorising a list trains. A list gives you recognition of a word and no evidence at all about its grammar, its register, its usual collocates or its connotation. Recognition is what gets you a mark on a reading comprehension question. Production under time pressure is what these four traits measure, and they are not the same ability.
There is a second, blunter version of the same risk. Pearson's guides state that a response which does not deal properly with the prompt "due to significant amounts of pre-prepared/memorized material" is an irrelevant response and scores 0 on Content — after which, in Pearson's own scoring diagram, there is no further scoring at all. A bank of memorised "high-level" sentences dropped into an answer is exactly the material that rule describes. The long version of that argument is in why PTE templates do not work.
What any automated scorer has to do with your word choice
Pearson does not publish its algorithms and anyone claiming to know them is guessing. But the guide does name the technology, and you can reason from the format about what any automated scorer must do, which is what the rest of this section is. I am labelling it as reasoning, not as a claim about Pearson's internals.
The guide says the written portion is scored using the Intelligent Essay Assessor, powered by Pearson's Knowledge Analysis Technologies engine, using "a proprietary application of the mathematical approach known as Latent Semantic Analysis (LSA)" — described there as "an approach that generates semantic similarity of words and passages by analyzing large bodies of relevant text" (Academic guide, page 13).
Whatever the implementation, any system built that way has one structural property: a word's meaning is derived from the company it keeps in a large body of text. It has no access to what you intended. It has only what you produced and what words like yours normally sit next to.
Two consequences follow, and they hold for any distributional scorer, not just Pearson's:
A common word in its normal context is invisible. It matches the expected neighbourhood, contributes its meaning cleanly, and gives the scorer nothing to flag. That is precisely the state the descriptor calls "appropriate choice of words".
A rare word outside its normal context is doubly visible. It is an unusual token and it sits next to words it does not normally sit next to. A distributional similarity measure has no way to separate an ambitious word from a mismatched one — which is why the rubric puts the credit for breadth in a separate, capped trait rather than in the same judgement.
This is why "use more advanced vocabulary" is bad advice as given. The advice that survives contact with the marking is narrower: use the most precise word you actually control. Precision is rewarded in the descriptors; obscurity is not mentioned in any of them.
The blanks are a collocation test, not a vocabulary test
Now the second mechanism, and the one with the most practice value in it, because it is where most candidates think vocabulary size is being measured and it is not.
PTE Academic has three fill-in-the-blanks types (Academic guide pages 40 and 44):
| Type | Section | No. | What you do | Scoring |
|---|---|---|---|---|
| Fill in the Blanks (Dropdown) | Reading | 5–6 | Choose from a drop-down at each gap | +1 per correct blank |
| Fill in the Blanks (Drag and Drop) | Reading | 4–5 | Drag words from a bank into the gaps | +1 per correct blank |
| Fill in the Blanks (Type In) | Listening | 2–3 | Type the word you heard into the transcript gap | +1 per word spelled correctly |
PTE Core has the same three, with the reading drop-down type renamed Reading and Writing: Fill in the Blanks and scored for both reading and writing (Core guide page 35). None of the six carries a trait. (Pearson's PTE Academic format page still calls the Academic drop-down "Reading and Writing: Fill in the blanks" in its opening paragraph, but the Academic Score Guide and that same page's own note on the question both score it as reading only, so treat the older label as stale.)
Look at what the two reading types have in common. The word is given to you. In the drop-down, the correct answer is one of the options in front of you. In drag-and-drop, Pearson's own reading format page states there are "more words than gaps so you will not use all the words provided" — the answers are all in the bank, sitting beside distractors.
You are not being asked to produce vocabulary. You are being asked which of the words in front of you the slot will accept. And a slot in English accepts words on three grounds, usually in this order of how fast they settle the question:
- Grammar. The slot needs a noun, or a past participle, or an adjective. Anything of the wrong class is out before you read the sentence for meaning.
- Complement pattern. The words after the gap constrain it. A verb followed by to must be one that takes to. A verb followed by on must take on.
- Collocation. Among the survivors, English simply prefers one. You draw a conclusion; you do not make one.
Only after all three do you need meaning at all, and by then there is often one option left.
The listening type is genuinely different and it is worth being straight about it: there you type the word you actually heard, so it is a transcription and spelling item first. Collocation helps you recover a word the audio blurred — it does not choose one for you. Spelling is where those points are won and lost, and that has its own mechanics in PTE spelling.
The drill: predict the slot, not the word
This is the transferable skill, and it does not require PTE material to practise. It is also the one drill that transfers to the vocabulary traits, because judging what a slot accepts is judging appropriacy.
Take any 150 words of academic prose — a journal abstract, a science news article, a textbook paragraph. Then:
- Cover one content word per sentence. A verb, a noun or an adjective. Not the, not of.
- Before uncovering it, write down three things. What part of speech the slot needs. What the words after the gap force (a preposition, a that-clause, a singular or plural noun). Two words that would fit.
- Uncover, and score yourself on the first two only. Whether you guessed the author's exact word is irrelevant — often several words would have been correct.
- Log the slots where you got the constraint wrong. Those are your real gaps, and they are invisible if you only track right and wrong.
Ten minutes a day. What you are building is not a bigger vocabulary; it is the reflex of reading a gap forwards and backwards before choosing.
Here is the drill worked through, to show what "getting the constraint right" looks like:
| Sentence with gap | What the slot forces | Why most options die |
|---|---|---|
| Researchers have long ______ the decline to habitat loss. | Past participle taking to | attributed, ascribed, linked, traced survive. blamed takes on. caused takes no preposition. |
| The findings ______ doubt on the earlier model. | Verb that collocates with doubt on | cast is the default; throw is attested in British English. Put, place and give are not. Collocation, not logic. |
| Rising costs ______ from a shortage of raw materials. | Verb taking from, meaning "are caused by" | result, stem, arise, derive. Note result in means the opposite direction — one preposition flips causation. |
| A significant ______ in temperature was recorded. | Countable noun after a significant | rise, increase, drop. Not risen, not rising, not increasingly. |
| The policy is ______ on continued funding. | Adjective taking on | contingent, dependent, predicated. Reliant takes on too. Subject takes to. |
| The study ______ a strong correlation between the two variables. | Verb taking a direct object, reporting a finding | found, revealed, identified, reported, documented. Not told, not said. |
Notice what settled five of those six: grammar and the preposition. Not how many words the reader knew.
Words worth owning, grouped by the task they feed
You asked for a list, so here is one — with the honest caveat that comes with every list on this subject, stated in full at the end. These are grouped by semantic field and by the question type they can actually earn on, because a word that helps your essay may be worth nothing on Describe Image.
For Summarize Written Text and Summarize Spoken Text: reporting verbs
These two tasks are compression tasks. On PTE Academic, Summarize Written Text scores its Form point only if the answer is one single complete sentence and is not under 5 or over 75 words (guide page 33); PTE Core drops the single-sentence rule and bands Form on length instead, with the full two points at 25 to 50 words (page 31). Summarize Spoken Text takes its full Form mark at 50 to 70 words on Academic (page 42) and 20 to 30 on Core. A reporting verb that carries the author's stance in one word is the highest-value item you can own here, because it buys you sentence length back.
Learn them with their pattern, not alone. The pattern is where the marks are.
| Verb | Pattern | Example |
|---|---|---|
| attribute | attribute X to Y | attributes the decline to overfishing |
| argue / contend / maintain | + that clause | argues that the evidence is inconclusive |
| demonstrate / establish | + that, or + noun | establishes a link between the two |
| challenge / dispute / question | + noun | challenges the prevailing account |
| account for | + noun | accounts for most of the variation |
| stem from / arise from | + noun | stems from a policy change |
| result in | + noun (X causes Y) | results in higher costs |
| result from | + noun (X is caused by Y) | results from higher costs |
| highlight / underline | + noun | highlights an unresolved tension |
| outline / trace | + noun | traces the development of the technique |
The result in / result from pair is the whole argument of this article in miniature. Same verb, both perfectly ordinary, and the preposition decides which way the causation runs. Get it backwards and you have not shown a narrow vocabulary — you have misrepresented the source, which lands on Content before it ever touches Vocabulary.
For Write Essay and Write Email: hedging and stance
Both tasks are scored on Content first, and Content descriptors reward dealing with the prompt accurately. Overstating a qualified claim is a content error. These are the words that let you be precise about how strong a claim is.
| Field | Words |
|---|---|
| Likelihood | tends to, is likely to, appears to, may well, is unlikely to |
| Degree | largely, broadly, in part, to some extent, marginally, substantially |
| Concession | admittedly, granted, while it is true that, on balance |
| Limitation | with the exception of, in most cases, under certain conditions |
| Attribution | according to, on this view, proponents argue, critics counter |
For Write Essay and Describe Image: change, quantity and comparison
Describe Image has no vocabulary trait — but its Content descriptor at band 6 reads "A variety of expressions and vocabulary are used appropriately and with ease and precision throughout the response", at band 5 "A variety of expressions and vocabulary are used appropriately throughout the response", and at band 3 "The range of expression and vocabulary is narrow and simple expressions are used repeatedly" (Academic guide, pages 18–19). So vocabulary is folded into Content on that task rather than scored separately. The word that does the work at the top band is, again, appropriately.
| Field | Words | The constraint that catches people |
|---|---|---|
| Upward | rise, climb, surge, soar, accelerate | a sharp rise (noun) vs rose sharply (adverb) |
| Downward | fall, decline, drop, dip, contract | a decline in sales (what fell) vs a decline of 20 per cent (by how much) |
| Stability | level off, plateau, hold steady, stabilise | level off is intransitive — nothing follows it |
| Volatility | fluctuate, oscillate, vary | fluctuate between X and Y |
| Extremes | peak, trough, bottom out, hit a low | peak at a figure |
| Comparison | outstrip, lag behind, keep pace with, converge | lag behind takes an object; converge takes on or with |
| Quantity | the number of (countable), the amount of (uncountable), the proportion of, the share of | the number of students, the amount of funding — never swapped |
For Write Email on PTE Core: register triples
PTE Core's Write Email Vocabulary descriptor says it in different words from the others, and the top band is the revealing part (Core guide page 33):
2 Good command of lexis appropriate to the context of the given situation 1 Limited range of lexis. Some of the lexis is appropriate to the context of the given situation, but lexical shortcomings lead to some imprecision
"Appropriate to the context of the given situation." The prompt names your recipient and the reason for writing. Choosing a formal word for an informal recipient is not a safe default — it is a mismatch with the situation, which is what the descriptor measures.
Learn these as triples rather than as "advanced" replacements:
| Formal | Neutral | Informal |
|---|---|---|
| I am writing to request | I would like to ask for | Could I ask for |
| I would be grateful if you could | Please could you | Can you |
| I apologise for | I am sorry about | Sorry about |
| Please be advised that | I would like to inform you that | Just so you know |
| with regard to | regarding | about |
| at your earliest convenience | as soon as you can | soon |
| I look forward to hearing from you | Hope to hear from you | Let me know |
Pick the column the prompt asks for, then stay in it. Mixing columns inside one email is the most common way this trait drops from 2 to 1.
For Answer Short Question: the only task where the item is the vocabulary
Answer Short Question is the one question type where Vocabulary is the entire score: one point, appropriate word choice or not. It is also, on PTE Academic, a listening item that does not touch your speaking score at all. The practice items and the reason candidates lose questions they knew the answer to are in PTE short answer questions — the failure mode there is the ten-second window, not the word.
What no list can do, including this one
Pearson does not publish the item bank, and it rotates. Every vocabulary list for PTE — this one included — is a sample of the kind of language the test uses, assembled from published rubrics and from what test takers remember. It is not coverage, and a longer list is not closer to coverage. It is just longer.
More importantly, none of the four traits above scores recall. They score use. You can recognise ubiquitous on a flashcard and still lose the Vocabulary point by writing ubiquitous of instead of ubiquitous in, or by using it where common was the accurate word and ubiquitous overstates the claim.
So the test of whether a word is yours is not whether you can define it. It is:
- Can you say what part of speech it is without thinking?
- Do you know what preposition or complement follows it?
- Can you name two words it normally sits next to?
- Do you know whether it is neutral, formal or loaded?
If any answer is no, that word is a liability in a scored answer and an asset in a reading passage. Keep it for reading. Use the word you can answer all four questions about.
And the practical ranking, if you have limited weeks: collocation and complement patterns first, precision second, breadth last. That is the order the descriptors put them in, and it is exactly the reverse of the order a word list puts them in.
Find out which trait your words are actually costing you
None of this is actionable until you know whether vocabulary is where you are losing marks at all — and on a real PTE score report, you cannot tell. Pearson removed the Enabling Skills section from the score report in November 2021 and replaced it with a Skills Profile. What a test taker receives today is one overall score and four communicative skill scores. The traits still decide your marks; you simply never see them.
That is the gap a preparation platform exists to fill.
Hilingo runs full PTE Academic and PTE Core mocks under real timing, and your first full scored mock is free with no card. There are 45 full PTE Academic papers and 45 section tests per section — a section test is one section of a full paper, offered on its own — and 30 and 30 for PTE Core. On a paid plan you can retake them as often as you like; the free tier is capped.
Writing and speaking answers come back marked per criterion, against the trait names and band descriptors Pearson publishes, rather than as a single number, so you can see whether a Summarize Written Text lost its point on Vocabulary, Grammar, Form or Content. Objective answers that carry a stored explanation come back with the reason the answer was wrong, which is what turns a missed drop-down blank into a collocation you now know. Read Aloud answers come back with pronunciation broken down to individual sounds in phonetic notation, plus pause count, pace, fluency and transcript. Any question, passage or transcript can be displayed in 50 languages if the English of the prompt is getting in the way of the English being tested.
We do not run Pearson's scoring engine — nobody outside Pearson does, and any platform claiming otherwise is selling you something. What we can do is mark against the published criteria, trait by trait, and show you the working.
- PTE marks distribution — the full task-to-skill map for both tests, with page numbers
- PTE short answer questions — the one task where Vocabulary is the whole score
- PTE spelling — the other trait that is marked but never reported
- Why PTE templates do not work — the Content-zero gate that memorised language triggers
- PTE Academic vs PTE Core — check which test's descriptors apply to you
- PTE score chart — PTE to IELTS and CEFR equivalents
Frequently asked questions
How many words do I need to know for PTE?
There is no number, and any page giving you one has invented it. Vocabulary is a named trait on four PTE question types and on three of the four the top band is awarded for appropriate word choice — a standard about accuracy, not about size. The practical answer is that the words you can use correctly under time pressure are the ones that score; the words you merely recognise help you on reading and listening comprehension and nowhere else.
Does using difficult words increase your PTE score?
Not on three of the four vocabulary traits, where the maximum band is awarded for appropriate word choice and there is no higher band to reach. On the fourth, Write Essay's Vocabulary Range, breadth is rewarded — but it is worth two of the twenty-six trait points on a task that appears once, and the six-point General Linguistic Range trait beside it requires that errors be "rare and minor". A difficult word used wrongly is a lexical error, which costs you the trait it was meant to win.
Which PTE tasks are scored on vocabulary?
On PTE Academic: Answer Short Question (Vocabulary 0–1), Summarize Written Text (0–2), Write Essay (Vocabulary Range 0–2) and Summarize Spoken Text (0–2). On PTE Core: Answer Short Question, Summarize Written Text, Write Email and Summarize Spoken Text. Every reading question type and every objective listening question type is listed as "Not applicable" under Traits scored, and every speaking task is scored on speaking traits instead — Content, Pronunciation and Oral Fluency on PTE Academic, with Appropriacy in place of Content on PTE Core's Respond to a Situation.
Is there a vocabulary score on the PTE score report?
No. Pearson removed the Enabling Skills section — which included a Vocabulary score — from the score report in November 2021 and replaced it with the Skills Profile. A current report shows one overall score and four communicative skill scores. The Vocabulary trait still exists inside the marking of individual tasks; it is simply no longer reported to you, which makes it harder to diagnose rather than less important.
How do I improve vocabulary for PTE fill in the blanks?
Mostly by not treating it as a vocabulary problem. On both reading fill-in-the-blanks types the correct words are in front of you — in the drop-down, or in a bank that Pearson says contains more words than gaps. What decides the blank is the grammar of the slot, the preposition or complement the following words force, and which of the survivors English actually collocates with. Practise by covering one content word per sentence in any academic paragraph and predicting the constraints on the slot before you look at the answer.
How can I improve my PTE vocabulary in two weeks?
Drop the list and work on patterns. Spend the time on complement patterns for twenty or thirty reporting and causal verbs you already half-know (attribute to, result in versus result from, stem from, account for), on the countable and uncountable pairs that show up constantly in academic prose (the number of versus the amount of), and on the daily slot-prediction drill above. Two weeks is not enough to widen a lexicon, and it is easily enough to stop making the errors that drop a Vocabulary trait from 2 to 1.