Search for a CELPIP-to-IELTS conversion and you will find a dozen confident tables on prep-company sites, all slightly different from one another, none of them citing a source. Meanwhile the two test makers jointly commissioned a real concordance study, had it analysed independently, and published it on both their own websites in 2025 — and almost nobody links to it.
The table below is the real one. What follows is where it came from, and the four things its own authors say about how far you can trust it.
The failure
The mistake this page prevents is treating a single line — "IELTS 7 = CELPIP 9" — as a fact about your own score report. Nobody publishing that line says where it came from, and the study that could have told them says something more complicated. The second mistake is assuming a concordance is an equivalency. It is not, and for an immigration application the difference matters.
Where the table comes from
The citation is:
Ikeda, N., Chen, M., & Clark, T. (2025). A concordance study of CELPIP General and IELTS General Training. Prometric.
The research was jointly conducted by Prometric and the IELTS Partners: British Council, IDP: IELTS Australia, and Cambridge University Press & Assessment. The analysis was not run in-house by either side — "The Australian Council for Educational Research (ACER) was commissioned to lead the analysis".
It is published on both sites. ielts.org published it 13 June 2025; the celpip.ca page shows published 14 May 2025, last modified 8 July 2026.
- https://www.celpip.ca/celpip-ielts-concordance-report/
- https://ielts.org/researchers/our-research/research-reports/a-concordance-study-of-celpip-general-and-ielts-general-training
That provenance is why this table is worth more than the ones on prep sites — and why the caveats below carry weight. The people who built it wrote them.
The table, and the warnings that come with it
Read as: the minimum CELPIP level that concords to each IELTS band.
| IELTS band | Reading | Listening | Speaking | Writing | Overall |
|---|---|---|---|---|---|
| 4.0 | 4 | 4 | 3 | 3 | 4 |
| 4.5 | 4 | 5 | 4 | 4 | 5 |
| 5.0 | 5 | 5 | 5 | 5 | 5 |
| 5.5 | 6 | 6 | 6 | 6 | 6 |
| 6.0 | 7 | 7 | 7 | 7 | 7 |
| 7.0 | 8 | 9 | 8 | 10 | 9 |
| 8.0 | 10 | 10 | 10 | 12 | 11 |
The report's own cautions, verbatim, in the same breath as the numbers:
"We advise caution when interpreting the results, particularly against using the concordance tables in isolation."
"Score users… are advised to interpret the concordance tables with caution and to consider supplementary evidence to support their decisions."
"Differences in test-taker populations and small sample sizes at extreme score levels may lead to imprecise concordance results."
That third one is worth sitting with. The rows at the top and bottom of the table — the ones people are most likely to look up — are the rows the authors flag as built on the least data.
The four skills do not move together
This is the most useful observation on the page, and it is visible in a single row.
At IELTS 7.0, the concorded CELPIP levels are Reading 8, Listening 9, Speaking 8, and Writing 10. Four different answers, spread across three levels, for one band.
Compare the rows below it. At 5.5 and 6.0 the four skills sit flat — 6, 6, 6, 6 and 7, 7, 7, 7. The relationship is not a constant offset. It changes shape as you go up, and it changes differently for each skill. Writing pulls away hardest: by IELTS 8.0 it concords to CELPIP 12 while the other three concord to 10.
Any page selling you a single "IELTS 7 = CELPIP 9" line is flattening something the actual study says is not flat. Nine is the overall figure in that row. Three skills sit below it, one sits well above it. Carry the overall number into a per-skill decision and you are using a number the table does not support.
What this study is not
It is not an IRCC document, and it does not behave like one.
The report carries no IRCC endorsement, and IRCC does not reference it. IRCC maps each designated test to the Canadian Language Benchmarks separately — CELPIP has its own equivalency chart, IELTS has its own, and CLB is the only common currency IRCC recognises. There is no official route from a CELPIP level to an IELTS band inside the immigration system, because the system never asks that question. It asks what CLB level you have.
So: for an immigration application, the concordance table is interesting, not operative. The CLB charts are the ones that decide anything. Check IRCC's own equivalency pages before you act on any number, including the ones on this page.
One thing the report leaves genuinely unclear
The report carries an immigration disclaimer, and the disclaimer names Australia, not Canada. It says the scores do not replace tests accepted for Australian visa purposes.
Whether the authors intended that to cover Canada is not clear from the document. We are not going to tell you it does, and we are not going to tell you it does not — it is an ambiguity in the source, and the honest thing is to report it and stop. Either way, the point above stands: IRCC works in CLB.
What actually costs you
- Quoting the "Overall" column at someone who needs a per-skill number. The overall figure hides a three-level spread at IELTS 7.0.
- Treating a concordance as an equivalency. A concordance is a statistical relationship between two score distributions. An equivalency chart is an administrative rule. Only one assesses applications.
- Reading the extreme rows as precise. The authors specifically warn that small samples at the extremes may produce imprecise results.
- Trusting an uncited table. If a conversion table does not name Ikeda, Chen & Clark or link to celpip.ca or ielts.org, you have no idea where its numbers came from.
How to actually use it
Use it as a rough sense of where you stand. If you have an old IELTS General Training result and you are weighing up CELPIP, the table tells you roughly which part of the CELPIP scale your past performance sat in — skill by skill, not as one number. That is a genuinely useful thing to know before you start preparing.
What it will never be is a substitute for sitting the test you actually need. A concorded level is not a score. It appears on no report, it can be submitted nowhere, and nobody assessing your application will accept it in place of a result. If a program requires a specific test, take that test.
Try one
[Try a marked task →] — one full task, marked, no card required.