Pro Membership
Unlock all premium features
The CELPIP Writing test runs 53 minutes and contains two tasks. Task 1 is an email of about 150 to 200 words in 27 minutes. Task 2 is a survey response of about 150 to 200 words in 26 minutes. Both tasks carry equal weight. Four criteria set your score: Content and Coherence, Vocabulary, Readability, and Task Fulfillment.
Reviewed September 2026
Scoring CLB 9 (Canadian Language Benchmark level 9) or higher on CELPIP Writing means satisfying an automated scoring system and a certified human rater inside one 53 minute exam. Paragon Testing runs every response through both stages before assigning a final score, and IRCC (Immigration, Refugees and Citizenship Canada) converts that score into CRS (Comprehensive Ranking System) points. Strong English is not enough on its own. You can write flawless grammar and still lose points in Task Fulfillment by skipping a single required bullet.
Two timed tasks fill the 53 minute module, and both carry equal weight in the final score. You write the email first, in 27 minutes, then the survey response immediately after, in 26 minutes. Neither task can rescue the other. A polished email paired with a rushed, unfinished survey response still produces a lower overall band than two solid, complete answers. The full section breakdown lives in the CELPIP Writing section format guide.
You get 27 minutes to plan, write, and check one email of about 150 to 200 words, built around a practical everyday situation. A simple split protects that window: four to five minutes to outline, 15 to 18 minutes to write, then three to five minutes to proofread. Skip the outline and you will spend your proofreading time fixing disorganised ideas instead of catching typos.
Structure decides more of your score than vocabulary does. Start with an appropriate greeting, state your purpose in the opening, develop every required point in the body, and close with a suitable sign-off. Tone shifts with the audience, so a complaint to a landlord needs formal wording while a note to a friend allows relaxed phrasing. The CELPIP Writing Task 1 email template shows how professional and casual structures diverge paragraph by paragraph, and there are more than 80 Task 1 email prompts here to practice on.
26 minutes go to the survey task, and the first two should decide which side of the argument you take. Arguing both sides triggers a penalty, because the rubric rewards one clear, well-supported position rather than a balanced discussion. A five-part sequence keeps it easy for both graders to follow: state your opinion, give a first reason, support it with an example, give a second reason, then close by restating your position. The CELPIP Writing Task 2 survey template shows how this structure adapts across unrelated prompts, and more than 74 Task 2 survey prompts are available to practice on.
Every response is typed directly into a locked browser interface. The countdown timer starts when you click begin and cannot be paused on either task. You navigate between prompts using the built-in interface and submit each response electronically before time runs out.
The stated target is 150 to 200 words, and a buffer of roughly 10% extends the real acceptable range to about 135 to 220 words. A 138 word response still clears the length check. Falling under 140 usually means a required detail got cut, and that missing detail is what costs you points in Task Fulfillment. Writing past 220 words carries a different risk: it leaves too little time to proofread, which is where extra spelling and grammar errors slip through.
The screen includes a basic spell checker that underlines misspelled words in red, and right-clicking a flagged word brings up suggestions. That tool catches typos and nothing more. Wrong verb tenses, awkward sentence structure, and weak organisation all pass through untouched, so every candidate targeting CLB 9 or higher has to catch those manually while proofreading. The rules around the spell checker and retakes are covered in CELPIP Writing spell check and retake rules.
An AI system reads every response first, measuring sentence variety, grammatical accuracy, and vocabulary range within seconds. A certified human rater then reviews the same text, judging whether it answers the prompt completely, holds an appropriate tone, and reads naturally from start to finish. Your final CLB score reflects both stages combined, not the judgment of a single grader.
Strong grammar satisfies the automated pass but does not guarantee a strong score, because the human rater checks qualities software cannot measure reliably: tone, logical flow, and whether the response actually answers what was asked. A response with three grammar mistakes that addresses every bullet point typically scores higher than a grammatically clean one that skips a requirement. CELTESTPIP gives you the same style of instant feedback. Submit a response and you get an estimated CLB score in seconds, broken out across Coherence, Vocabulary, Readability, and Fulfillment.
Four categories carry equal weight in the final CLB level, and strength in one does not offset weakness in another.
A criterion-by-criterion breakdown of what raters look for sits in how CELPIP Writing is scored.
Almost always, the missing bullet. Not the grammar.
Take the free Task 1 prompt on this page. You are planning a week-long trip to Halifax to visit your elderly parents. You have a dog named Biscuit who cannot come with you. Your neighbour has met Biscuit several times and looked after him for a weekend last summer. Write an email to that neighbour in about 150 to 200 words, doing three things:
Here is what happens in practice. Candidates with genuinely good English write a warm, fluent, well-organised email covering the trip and the dog's routine in detail, then run out of words and close politely without ever offering anything in return. The grammar is clean. The tone is right. The third bullet never got answered, so Task Fulfillment drops, Content and Coherence follows it down, and a response that reads like CLB 9 English scores CLB 7.
The fix costs you one sentence. "If it helps, I am happy to water your plants while you are away in September, or return the favour whenever you next travel." That is the whole difference, and it is why counting the bullets before you start typing is worth more than any vocabulary list.
The free prompt gives you three graded sample responses to compare against, labelled Basic, Good, and Excellent, plus a scored breakdown showing where each one gained and lost points. Try it without logging in.
CELPIP Writing is rated on the CLB scale from 1 to 12 rather than a raw question count. The table below describes what each band demonstrates across the AI and human stages, and the CRS immigration points it earns per skill.
| CLB Level | What the AI and human raters look for | CRS points per ability |
|---|---|---|
| 11 or 12 | Sophisticated vocabulary, refined organisation, near error-free grammar, full nuanced task completion | 34 |
| 10 | Strong control, varied vocabulary, only minor slips | 34 |
| 9 | Clear organisation, varied vocabulary, accurate grammar, full task coverage | 31 |
| 8 | Ideas developed competently, occasional errors that do not block meaning | 23 |
| 7 | Ideas addressed but basic development, functional word choice, minor grammar slips | 17 |
| 6 | Simple, repetitive vocabulary, errors that sometimes affect clarity | 9 |
| 4 or 5 | Basic everyday English, limited range, frequent errors | 6 |
Band descriptions summarise the official CELPIP Writing rubric of Content and Coherence, Vocabulary, Readability, and Task Fulfillment. The CRS figures are points per ability for an applicant without a spouse or common-law partner, first official language, published by IRCC on the Comprehensive Ranking System criteria page. Applicants with an accompanying spouse or common-law partner earn slightly fewer points per ability: 32 at CLB 10 or more, 29 at CLB 9, and 22 at CLB 8. Express Entry also requires CLB 7 in all four abilities for the Federal Skilled Worker Program and for Canadian Experience Class applicants in TEER 0 and TEER 1 jobs, and your results must be less than two years old when you apply.
For the full cross-reference to IELTS and TOEFL and how each level reads, see the CELPIP Writing score chart.
Memorized paragraphs cost you points because both graders are built to flag rigid, unnatural phrasing. A human rater can spot a pasted introduction within the first two sentences, since it rarely connects cleanly to the specific prompt on screen. Picture the common version of this: a polished, generic customer-complaint paragraph dropped onto a prompt about a delayed transit refund. The grammar is fine, but the mismatch reads as off-topic, and Task Fulfillment and Coherence both take the hit.
Memorize the sequence, not fixed paragraphs. For Task 1: greeting, purpose, supporting details, closing. For Task 2: opinion, reason, example, reason, conclusion. Keep the framework in mind and build fresh sentences around what the prompt actually asks. That single habit is what separates a CLB 9 response from a CLB 7 one.
Practice the sequence under real conditions with the CELPIP Writing course and the timed prompts on this page. When you are ready to sit the whole thing end to end, the full timed mock exams run Writing back to back with the other three sections.
Start with the free email prompt or the free survey prompt. If you want to practice across all four sections rather than Writing alone, the full question bank covers every task type on the test.
Two computerised tasks fill 53 minutes: an email of about 150 to 200 words in 27 minutes, followed immediately by a 26 minute survey response that requires picking one side of an opinion-based prompt. Both tasks contribute equally to the final score.
Paragon grades every response through an AI-Human hybrid system built on four equal categories: Content and Coherence, Vocabulary, Readability, and Task Fulfillment. The algorithm checks grammar and sentence structure first, a human rater then verifies tone and prompt alignment, and the combined result lands between CLB 1 and CLB 12.
Two, one in each task, with no account required. The free item is always the first in each list and includes the timer, the graded sample responses, and the AI score breakdown.
Usually a missed instruction rather than weaker English. Task 1 prompts carry three explicit bullets and all three must be answered. A fluent, well-organised email that quietly drops the third bullet loses points in Task Fulfillment and Content and Coherence, which is enough to pull a CLB 9 level of English down to a CLB 7 score.
CELTESTPIP scores writing samples using AI within seconds and returns an estimated CLB score immediately, broken out across Coherence, Vocabulary, Readability, and Fulfillment. The first prompt in each task is free to try without an account.
Answer every bullet in the prompt using the correct tone for the situation. A clear structure helps: greeting, purpose statement, organised body, proper sign-off, all within 220 words. Planning for four to five minutes before typing is the single habit that separates CLB 9 responses from CLB 7 ones.
Task 1 asks for correspondence, Task 2 asks for a defended opinion. The email solves a specific problem for a named recipient. The survey response picks one side of a community issue and supports it with concrete examples, never both sides at once.
Reduce repetitive vocabulary and grammar slips, then expand your supporting examples until every prompt requirement has a direct answer. A CLB 6 response shows simple, developing English with errors that sometimes block meaning. CLB 7 requires functional word choices that fully address the prompt without confusing the human rater.
Yes. CELTESTPIP runs a countdown on both Writing tasks using the real exam limits of 27 minutes for Task 1 and 26 for Task 2, and the full mock exams run them back to back so you practice the pacing under exam-day conditions.
Writing under 140 words costs you points immediately in Task Fulfillment. Memorized templates cost points too, since both the AI pass and the human rater catch pasted paragraphs quickly. Arguing both sides of a Task 2 survey instead of holding one clear position is the single most common reason strong writers still miss CLB 9.
About the author
