Almost everyone asking this wants the same thing. You want CLB 9, because CLB 9 is where the Comprehensive Ranking System starts paying properly, and you want the test that gets you there with the least work.
Two facts decide it. Canada does not read your test score, it reads a Canadian Language Benchmark, and the conversion chart is different for each test. Your CLB is worked out per skill, so the level you can claim is the lowest of your four. The test that suits you is the one whose chart asks least of your weakest skill.
One correction before the tables, because this one costs people a test fee. For Express Entry, Immigration, Refugees and Citizenship Canada accepts IELTS General Training and PTE Core. Not IELTS Academic. Not PTE Academic. The IRCC language requirements page lists PTE Core as the accepted Pearson option, and there is no PTE Academic to CLB conversion for economic immigration. A PTE Academic result is unusable in an Express Entry profile however high it is.
BandNine is an IELTS practice platform built by two teachers. We do not sell PTE preparation, so where PTE is the better fit below, we have said so.
#What IELTS score do you need for CLB 9?
For CLB 9 on IELTS General Training you need Listening 8.0, Reading 7.0, Writing 7.0 and Speaking 7.0, and you need all four at once. Here is the full chart from CLB 7 to CLB 10, taken from the IRCC page that lists language results for post-graduation work permits and confirmed row for row against the IRCC page for the Rural and Northern Immigration Pilot.
| CLB | Listening | Reading | Writing | Speaking |
|---|---|---|---|---|
| 10 | 8.5 | 8.0 | 7.5 | 7.5 |
| 9 | 8.0 | 7.0 | 7.0 | 7.0 |
| 8 | 7.5 | 6.5 | 6.5 | 6.5 |
| 7 | 6.0 | 6.0 | 6.0 | 6.0 |
Read the Listening column on its own. To reach CLB 9 you need Listening 8.0, a full band above the Reading, Writing and Speaking thresholds on the same row. Between CLB 7 and CLB 8 it jumps a full band and a half while every other skill moves half a band. Check any third-party table you find against this row first. If a chart shows CLB 8 Listening at 6.5 or 7.0, that chart is wrong.
The top row is not symmetrical either. At CLB 10 you need Reading 8.0 but Listening 8.5, while Writing and Speaking need only 7.5. CLB 7 is the one row where all four skills sit at the same band.
#What PTE Core score do you need for the same benchmarks?
For CLB 9 on PTE Core you need Listening 82 to 88, Reading 78 to 87, Writing 88 to 89 and Speaking 84 to 88. The same two IRCC pages carry this chart, again identical.
| CLB | Listening | Reading | Writing | Speaking |
|---|---|---|---|---|
| 10 | 89 to 90 | 88 to 90 | 90 | 89 to 90 |
| 9 | 82 to 88 | 78 to 87 | 88 to 89 | 84 to 88 |
| 8 | 71 to 81 | 69 to 77 | 79 to 87 | 76 to 83 |
| 7 | 60 to 70 | 60 to 68 | 69 to 78 | 68 to 75 |
Two things to notice. Writing sits higher up the PTE scale than the other skills at every level, so on PTE it is Writing that squeezes you at the top. And at CLB 10, PTE Writing is 90 exactly, a single value rather than a range.
Do not lay the two tables side by side and decide which looks kinder. The scales are unrelated, and a band and a point are not comparable units. Each table tells you the demand on each of your skills, and what that costs in study hours is a question about you.
#What is CLB 9 worth in CRS points compared with CLB 7?
Moving from CLB 7 to CLB 9 in all four abilities is worth 56 CRS points on the first official language if you have no accompanying spouse, and 52 points if you do. That is the real reason this question gets asked. IRCC awards these points per ability, not per test.
| Level in each ability | Points per ability, with spouse | Points per ability, without spouse |
|---|---|---|
| CLB 10 or more | 32 | 34 |
| CLB 9 | 29 | 31 |
| CLB 8 | 22 | 23 |
| CLB 7 | 16 | 17 |
Four abilities at 34 points is the published maximum of 136 without a spouse, and four at 32 is the published maximum of 128 with one, which is a useful check that you are reading the table correctly. Note the top row label. IRCC writes "CLB 10 or more", so there is no tier above it. A candidate at CLB 11 or CLB 12 scores what a candidate at CLB 10 scores.
The steps are lopsided. Without a spouse, an ability is worth 17 points at CLB 7 and 31 at CLB 9, so lifting one is 14 points and lifting all four is 56. With a spouse it is 16 to 29, so 13 per ability and 52 in total. The last step, CLB 9 to CLB 10 or more, is worth 3 points per ability, so 12 across all four either way. Big at the bottom, small at the top.
#The skill transferability points that are easy to miss
If you have foreign work experience, crossing CLB 9 pays a second time. Under skill transferability, CLB 7 or more in all four abilities with at least one below CLB 9 earns 13 points with one to two years of foreign work experience and 25 points with three years or more. CLB 9 or more in all four earns 25 and 50 points for the same two bands of experience.
Put the two together. A candidate without a spouse who has three or more years of foreign work experience and moves from CLB 7 to CLB 9 across the board gains 56 points on the first official language and another 25 on skill transferability. The whole core human capital section is capped at 500 points without a spouse and 460 with one.
Second official language points are much smaller, and the cap catches people out. Per ability it is 1 point at CLB 5 or 6, 3 points at CLB 7 or 8, and 6 points at CLB 9 or more, with a maximum of 22 with a spouse and 24 without. Six points across four abilities would be 24, so with a spouse the cap binds.
The floors matter too. The Federal Skilled Worker Program requires CLB 7 in all four abilities in your first official language. The Canadian Experience Class requires CLB 7 for NOC TEER 0 or 1 jobs and CLB 5 for TEER 2 or 3 jobs. Every figure in this section comes from the IRCC Comprehensive Ranking System criteria page and the Express Entry language test results page. Our fuller walkthrough sits on the Canada PR page.
#Which test is easier to sit?
Neither test is easier in general. What is true is that the two tests fail different people, and you can usually predict which one will fail you.
The structural difference is who marks you. IELTS Writing and Speaking are marked by trained human examiners, and IELTS Speaking is a live conversation with one of them. PTE Core is delivered and scored by computer throughout, including speaking. Listening and Reading are machine marked on both tests, so that half of the decision is mostly format preference.
#A person or a machine: which speaking format suits you?
A live examiner rewards interaction and a machine rewards consistency, and you probably already know which of those you are better at. Here is the case for each, as we would put it to a student.
The case for a person. An examiner adapts. If a question lands badly they rephrase it, and if you restart mid-sentence they hear the repair rather than a hesitation. A person also shows you they have followed you, and a microphone gives you nothing back, which changes how you sound as you go. If your fluency comes from having someone to talk to, an empty room and a microphone will cost you something.
The case for a machine. Every candidate gets the same delivery, in the same order, with the same timing. There is no examiner having a difficult afternoon, no unexpected follow-up, no small talk to misread. If interviews make you freeze, that alone can outweigh any difference in the conversion chart. The trade is that a scoring model weighs how you say it as much as what you say, so heavy hesitation, long pauses and very fast delivery cost you even when the content is good.
Both of those are teaching judgements, not measurements. Record yourself doing each format and see which version of you sounds better. Do that before you book.
#Should you choose IELTS or PTE?
Choose the test whose chart asks least of your weakest skill, then use format as the tie breaker. In practice that resolves as follows.
Choose IELTS General Training if:
- You speak more fluently to a person than into a microphone, or your delivery has features you would rather a human listener interpreted.
- You already have a recent IELTS result, so you know which of the four skills is holding your CLB down.
- You are also making applications that accept IELTS General Training, so one sitting can serve both. Check the version each one wants first, because most universities ask for Academic and an Academic result is no use to IRCC.
Choose PTE Core if:
- Interviews make you freeze, and you perform better when nobody is watching.
- You are comfortable at a screen under a timer, speaking to a prompt with no conversational partner.
- You want marking that is uniform by construction, with no human in the room.
Choose neither yet if the skill capping your CLB is weak on any format. Test shopping cannot fix a Writing 6.0. If Writing or Speaking is your floor, the work is the same on either test: write or speak, get it marked, act on the marking. Changing provider does not change that. Our free diagnostic scores one writing or one speaking sample, so it can show you where that skill sits before you book. It does not cover Listening or Reading, so for those two sit a timed practice test and mark it against the key.
If CELPIP is also on your list, we have a three-way comparison at IELTS vs PTE vs CELPIP for Canada, with a method for working the charts against your own profile.
#Which rules catch Canada-bound candidates out?
Four rules are the ones worth checking before you pay for anything, and all four are on IRCC's own pages. Read them before you book anything.
- Your weakest skill sets your level. CLB is per ability and CRS pays per ability, so one skill at CLB 7 while the others sit at CLB 9 does not give you a CLB 9 profile.
- IELTS One Skill Retake does not count. IRCC does not accept IELTS One Skill Retake results for Express Entry. If one skill lets you down, the whole test has to be sat again. More on that in our guide to One Skill Retake.
- Results must be less than two years old both when you complete your Express Entry profile and when you submit your application for permanent residence. A result that is valid on the day you create the profile can expire underneath you.
- Only the right version counts. IRCC's language requirements page lists CELPIP-General, IELTS General Training and PTE Core for English, and TEF Canada and TCF Canada for French. IELTS Academic and PTE Academic are not on it.
#Has anything changed for 2026?
No. As of today the conversion tables, the list of accepted tests and the CRS language points are all unchanged. The only development is a consultation IRCC has opened on reforms to Express Entry's federal high skilled programs and the CRS. It is a proposal document with no implementation dates, and it says IRCC is considering keeping the existing points for first and second official language proficiency. It floats two language ideas, neither in force: exempting candidates whose Canadian professional licensing already assessed their language, and reducing or removing the CRS bonus points for French. Nothing in it should change what you book this month.
One practical note. The IRCC page that used to hold the canonical equivalency charts is returning a 404 at the time of writing, and IRCC's own help centre still links to it. The live tables sit on the post-graduation work permit language results page and the Rural and Northern Immigration Pilot language testing page, which is where the charts above came from. Check whichever page loads for you before relying on any number, here or anywhere else.
#What to do this week
Start from your own four skill levels, because every part of this decision runs off them. If you do not have recent numbers, sit one timed practice in each skill. Put those bands into the band calculator and read the CLB your weakest skill supports, not the one your average suggests. Then find the CRS points that level earns in the table above, and the points the next level up would earn. That gap is the size of the prize.
Only after that should you choose a test. If you go the IELTS route, our writing feedback is deterministic: resubmit the same text and you get the same band and the same comments, not a fresh opinion. That is a consistency property, not an accuracy one. We publish the accuracy figure every Saturday at /quality with the method beside it. It covers Writing, it is measured against labels our two teachers write from the public band descriptors, and we do not publish a figure for Speaking, Listening or Reading.