CELPIP Vocabulary: The Words That Matter (and the List You Don't Need)

Somewhere right now, a candidate is memorizing a 500-word list of "advanced CELPIP vocabulary" — words like "commence" and "utilize" — and quietly lowering their score. Not because effort is bad, but because the effort is pointed the wrong way: CELPIP tests everyday functional English, and a misplaced impressive word costs more than a plain one earns.

This is the vocabulary post that tells you what actually moves scores: which words the test really draws from, which habits pay, and the acquisition method that makes words stick. If you want your current level measured first, the free mock covers all four skills → celpipuni.com/free-mock-test.

TL;DR

What matters in CELPIP vocabulary: (1) topic clusters for the everyday domains speaking and writing prompts live in; (2) the precision toolkit — collocations like "heavy traffic," not "strong traffic"; (3) speculation and attribution language ("probably," "apparently," "she argues") for speaking, listening, and reading viewpoints; (4) connectors used sparingly. The method: encounter words in context (podcasts, articles), capture five a day, force them into that day's practice task, review weekly. What hurts: memorizing decontextualized lists and deploying half-owned "fancy" words — the #1 band-capper we see in graded writing.

What CELPIP actually draws from

CELPIP's scenarios are deliberately everyday: neighborhoods, work, technology habits, health routines, small decisions. (The speaking topic map shows the actual prompt families.) So the vocabulary that pays isn't "academic word list" material — it's the natural, precise, everyday lexicon of:

city life: commute, bylaw, renovation, noise complaint, lease
work: shift, deadline, reschedule, follow up, work out (a plan)
daily decisions: weigh the options, worth it, trade-off, come down to
people and attitudes: laid-back, reliable, pushy, easy-going, down-to-earth

Notice what those have in common: they're the words you'd actually use describing your real life. Raters score vocabulary on range and precision — using the right natural word, not the fanciest one. "The landlord's been dragging his feet on the repair" scores; "the landlord has exhibited considerable procrastination" doesn't.

The four categories that matter

1. Topic clusters (for speaking and writing content)

Prompts cluster around a handful of everyday domains. For each domain, know its 15-20 core words cold — not 500 random words, but the working vocabulary of a domain: city/civic life, workplace, technology habits, health and routines, money and decisions, relationships and neighbors. (The topic bank organizes them by task.)

2. The precision toolkit (collocations)

Collocations are where intermediate speakers leak fluency: the individual words are right, the pairing is wrong. "Strong traffic" instead of "heavy traffic." "Make a decision" fine, "do a decision" wrong. "Take a break" vs "have a break" (both fine — but "make a break" means escape). Precision at this level is what separates a 7 from a 9 in both writing and speaking criteria.

The wrong version	The right version	Why
strong traffic	heavy traffic	adjective-noun pairing
make a research	do research / research it	verb-noun pairing
a good option of	a good alternative to	preposition fit
very good conditions for	ideal conditions for	natural intensifier
in the other hand	on the other hand	fixed phrase

3. Speculation and attribution language

Underrated and heavily tested: the words that carry uncertainty and stance. "Probably," "I'd guess," "it looks like," "apparently," "supposedly" — speaking tasks 3-4 (describing, predicting) live on these, and listening/viewpoints parts test whether you hear them (attribution skills). "She argues" vs "she notes" is the reading viewpoints part's core distinction.

4. Connectors — sparingly

Contrary to prep-culture folklore, connectors are seasoning, not structure. A response with "moreover, furthermore, additionally" every second sentence reads as template-stuffed — a band-limiter, not a band-raiser. The 9-level writer uses few connectors because their logic connects the ideas: "There's one more thing worth fixing first" beats "Furthermore."

The method: how words actually stick

The acquisition loop that works (and it's the same loop for listening and reading gains — the practice method):

Encounter in context. Read and listen to everyday North American material. Words met in context arrive pre-attached to meaning and situation — words from lists arrive pre-attached to nothing.
Capture five a day. Not twenty: five, chosen because you almost knew them — the words on the tip of your tongue. Those are one exposure from yours.
Force-use each same day. Each captured word goes into that day's practice — a speaking task, an email, an answer summary. Use is the consolidation step; capture without use is souvenir-collecting.
Weekly review, usage-recall style. Don't re-read definitions — try to produce each word in a new sentence. Production recall, not recognition.

The mistakes that cost bands

The thesaurus upgrade. The single most common vocabulary error in graded writing: replacing plain correct words with half-owned impressive ones. "Myriads of individuals" for "many people," "endeavor to" for "try." Raters read it instantly — imprecise in context, slightly off in register — and the vocabulary criterion drops while the writer believed they were upgrading. If you'd never say it in conversation, don't write it in a CELPIP email.

Memorizing without context. Word lists strip away everything that makes retrieval possible in real time: situation, collocation, grammar pattern. The result is vocabulary you can recognize in print and never produce under a 60-second clock.

Ignoring your own gaps. Your miss-logs (error taxonomy) tell you exactly which words you lacked and where. Captured from your logs, the five-a-day are bespoke; captured from a generic list, they're someone else's curriculum.

Mistaking vocabulary for the bottleneck. Many 7s aren't vocabulary-limited at all — they're precision- and delivery-limited (the 7-to-9 writing case study). Diagnose before you drill: if the transcript shows you knew every word, vocabulary isn't your project.

What good looks like

The 9-level sentence isn't the one with the rarest word — it's the one with the realest one. Compare, on the same task (advising a friend on a noisy neighbor):

A (thesaurus mode): "I would advise commencing with an amicable approach prior to escalating the matter to the pertinent authorities."

B (owned and precise): "Start friendly — a quick note or a knock on the door. Most people genuinely don't know how much sound carries in these buildings. If that gets you nowhere in a week or two, then it's worth looping in the property manager."

B scores higher on every criterion: vocabulary (precise, natural, idiomatic — "looping in," "gets you nowhere"), coherence (the logic connects without connectors), readability, task fulfillment. Not one rare word in it. That's the whole vocabulary lesson in two sentences: own real words cold rather than borrowing fancy ones lukewarm.

To see whether vocabulary is actually your gap — and which of the four categories is leaking — a graded evaluation of your writing and speaking samples answers it precisely → celpipuni.com/evaluation.

The nuance: vocabulary breadth is a listening problem too

Here's the connection candidates miss: vocabulary breadth isn't just about producing better sentences — it's the upstream cause of most listening misses. When the transcript review (the loop) shows "I caught sounds but not meaning," that's a vocabulary gap wearing a listening costume. And in listening, vocabulary work pays double: the word you captured from yesterday's podcast is the word you'll decode at speed in next month's test audio.

So the priority question is: your misses — are they production (writing/speaking precision) or comprehension (listening/reading coverage)? Same five-a-day habit either way, but pointed at different domains. Diagnose first, then drill.

Your checklist for this week

Five-a-day capture habit started — from your own practice material
Each day's five words forced into that day's practice task
One collocation audit: your last three writing tasks, highlight every "almost-right" pairing
Speculation language drilled once out loud (describe a photo: probably / looks like / I'd guess)
Weekly review by producing sentences, not re-reading lists

Tools and resources

CELPIP speaking topics — the prompt domains worth clustering vocabulary around.
CELPIP grammar: the rules that move scores — the accuracy side of the same coin.
Listening practice sources — where contextual encounter happens daily.
Our evaluation service → celpipuni.com/evaluation — pinpoints which vocabulary category is leaking.

FAQ

Is there an official CELPIP vocabulary list?

No — CELPIP doesn't publish one, and any third-party "CELPIP word list" is someone's product, not a syllabus. The test draws from everyday functional English; domain clusters (city life, work, routines, decisions) serve better than any list.

How many words do I need for CELPIP 9?

Fewer than you think, known better than you'd guess. A few thousand well-owned everyday words — including collocations and phrasal verbs — outperform double that in half-owned condition. Precision and naturalness are the scored dimensions, not rarity.

Should I memorize word lists for CELPIP?

Not as your main method. Lists strip context, collocation, and grammar patterns — the things retrieval under time pressure needs. Encounter-capture-use (five a day, forced into practice) builds producing vocabulary; lists build recognizing vocabulary.

Are phrasal verbs important for CELPIP?

Very — everyday North American English runs on them ("put up with," "run late," "sort out," "loop in"). They're also exactly what "fancy word" culture steers people away from. Phrasal verbs in natural use signal exactly the register CELPIP rewards.

What vocabulary helps the listening section?

Words captured from audio — met with the ears, not the eyes. The five-a-day habit run on podcast material pays double: production vocabulary and listening coverage at once (the loop).

How do I know if vocabulary is actually my weakness?

Diagnose before drilling: if transcripts show you knew every word you missed, your problem is strategy or attention, not vocabulary (the taxonomy). If graded writing keeps flagging precision errors rather than range gaps, you need collocation work, not new words.

What to do next

Start the five-a-day capture tomorrow — from your real practice material, not a list.
Force every captured word into same-day use.
Run one collocation audit on your recent writing this week.
Get one graded evaluation to confirm vocabulary is (or isn't) your bottleneck → celpipuni.com/evaluation

The bottom line

CELPIP vocabulary is five hundred owned words, not five thousand listed ones — real, precise, everyday, collocated, drilled through use. The thesaurus is where band 9s go to become band 7s.

Find out what your vocabulary is actually doing in graded work → Get Your Writing and Speaking Evaluated.