IGCSE Grade Boundary Myths: What Cambridge's Own Rulebook Says
Cambridge grade thresholds are not curved and not decided in advance — they track how demanding the paper was, measured against the same standard every year
TL;DR: The biggest IGCSE grade boundary myth is that grades are curved against other candidates. They are not. Across the last six sessions of Cambridge IGCSE Chemistry, the A* threshold has needed between 74.5% and 86.5% of the raw marks — and never once reached 90%. Every one of those candidates was still reported with a percentage uniform mark of 90 or above, which is exactly why so many students believe an A* requires 90% of the marks. Cambridge commits in writing that its examinations are not norm-referenced, so no fixed share of candidates is allocated each grade and another student's high mark cannot take your A* away.
Published 25 July 2026 · Last updated 25 July 2026
What this guide covers
- Are IGCSE grade boundaries curved?
- Where does the grade boundary curve myth come from?
- What an A* in IGCSE Chemistry really costs
- What does the number next to my grade mean?
- Does a PUM of 89 mean you beat 89% of candidates?
- What Cambridge weighs when it sets a threshold
- Are boundaries decided before the exam?
- Most of your grade boundaries were never judged by a human
- Why there is no A* on Paper 4
- One mark short — will they round you up?
- What if your 2026 exams were cancelled?
- Does Edexcel IAL work the same way?
- How to use past boundaries without fooling yourself
- About the author
- Frequently asked questions
Are IGCSE grade boundaries curved?
No. Cambridge states in its Code of Practice that its examinations will not be norm referenced, and it defines a norm-referenced assessment as one where the proportion of candidates awarded each grade is fixed regardless of changes in the ability of the cohort.
That single commitment kills the most damaging misconception in the whole topic. There is no quota of A*s. There is no rule that a set percentage must fail. If more candidates reach the required standard, more candidates can be awarded that grade — Cambridge commits that the proportion of successful candidates will increase when the knowledge, understanding and skills being assessed have improved over time.
There is one honest caveat in that same clause, and it is worth knowing: no such improvement in results will arise merely from candidates becoming more familiar with the nature of the assessment. Getting better at the subject moves results. Getting better at recognising the paper does not.
I hear this myth every single results season. Students tell me their class did well, so the boundaries must be about to rise. Parents forward messages from Facebook groups saying boundaries are going to be high this year. Neither is how the system works, and neither is worth the anxiety it causes.
Where does the grade boundary curve myth come from?
It comes from UK guidance being read as if it applied everywhere. Almost every page that explains grade boundaries is describing the UK system, not the one that awards your Cambridge IGCSE, and the two are not the same.
UK GCSEs and A levels are awarded using what Pearson calls the comparable outcomes approach: if the national cohort for a subject is similar in past-performance terms to last year, results should be similar nationally too. To do that, awarding bodies work with Ofqual to build a reference matrix for each subject, based on candidates' prior attainment — Key Stage 2 test results for GCSE, and GCSE results for AS and A level.
Now look at what that requires. A Key Stage 2 baseline. A national cohort. A regulator with jurisdiction over that cohort. A student sitting Cambridge IGCSE Chemistry in Kuwait has none of those things. The machinery that people describe when they tell you grades are "statistically controlled" is machinery that was never pointed at you.
This is where most of the confusion starts, and it starts with a detail almost nobody checks. Very few students and parents look at the exam code before they read something on social media. They see "GCSE grade boundaries" in a headline and assume every GCSE from every board in every country is the same qualification. It is not. Cambridge IGCSE Chemistry is 0620. Cambridge IGCSE Biology is 0610. Edexcel International A Level Biology units run WBI11 to WBI16. Those codes exist precisely because the qualifications are different, and they are graded by different rules.
For Cambridge specifically, the help centre puts it plainly: grade thresholds are not adjusted in line with the ability of students, but in line with the demand of the question paper in relation to the assessment criteria.
What an A* in IGCSE Chemistry really costs
An A* in Cambridge IGCSE Chemistry has needed between 74.5% and 86.5% of the available raw marks across the last six published sessions. It has not once required 90%.
Here is the data, taken from the official Cambridge threshold tables for Option CY — the standard extended combination of Paper 2, Paper 4 and Paper 6, weighted to a maximum of 200 marks. The final column is the percentage uniform mark that every candidate on that exact threshold received.
| Session | A* threshold (of 200) | Raw % needed | PUM reported |
|---|---|---|---|
| June 2023 | 157 | 78.5% | 90 |
| November 2023 | 149 | 74.5% | 90 |
| June 2024 | 170 | 85.0% | 90 |
| November 2024 | 158 | 79.0% | 90 |
| June 2025 | 173 | 86.5% | 90 |
| November 2025 | 153 | 76.5% | 90 |
Read the last two columns together, because that is the whole misconception in one line. Six different raw percentages. One reported number. A candidate on 149 out of 200 and a candidate on 173 out of 200 both got an A* and both saw 90 on their statement of results.
The June 2025 figure is verifiable in seconds if you want to check me: the official June 2025 Chemistry 0620 threshold table gives 173 for Option CY. My full session-by-session breakdown of all three papers, including why Chemistry thresholds swing far more than Biology, sits in the complete guide to IGCSE Chemistry grade boundaries.
What does the number next to my grade mean?
It is your percentage uniform mark, and Cambridge's guide to percentage uniform marks states directly that it is not the total mark achieved for the syllabus. It is a point on a common scale showing whether your performance sat close to the top, the middle or the bottom of your grade.
The mechanic is simpler than most explanations make it sound. The scale is fixed: A* is 90–100, A is 80–89, B is 70–79, C is 60–69, D is 50–59, E is 40–49, and for IGCSE only, F is 30–39 and G is 20–29. Your raw total is then mapped onto the band you landed in. Cambridge gives the worked example itself — a student on the minimum mark for A* gets a PUM of 90, and a student halfway between the D and C thresholds gets 55.
So a PUM of 85 means you were roughly halfway up the A band for that syllabus in that session. It does not mean you scored 85% of the marks.
Three practical details students miss. The PUM is stated at syllabus level only, so there is no PUM for an individual paper. It appears on your statement of results but never on your certificate. And no PUM is issued at all if your result is Ungraded, No Result, Pending or To Be Issued. One caveat on top of that: Cambridge supplies percentage uniform marks in countries where they appear on statements of results, so whether you see one depends on where you sat. If your slip shows only a grade, nothing has gone wrong.
Does a PUM of 89 mean you beat 89% of candidates?
No. A PUM describes your position inside a grade band, not your rank within the cohort — those are different quantities and the number can never tell you the second one.
A PUM of 89 means you were at the top of the A band, one point below the A* range. That is useful information: it tells you the A* was close. It tells you nothing whatsoever about how many candidates worldwide scored above you, because the scale is anchored to grade thresholds, not to percentiles.
This confusion runs deep in student communities. You can watch it play out in a Cambridge IGCSE forum thread asking whether 70 really is an A* — the question only makes sense if you have already assumed the number is a raw percentage or a ranking.
What Cambridge weighs when it sets a threshold
Cambridge commits to using a combination of expert judgements and statistical evidence, covering both the difficulty of the assessment and the ability of the cohort. Principal Examiners recommend thresholds — or ranges within which they believe thresholds should fall — at the end of the marking period, before grading begins.
Let me describe what that looks like from the examining side, because the published documents are accurate but bloodless. Thresholds are formed from performance across all candidates worldwide, weighed against how demanding the questions turned out to be. The relationship runs in the direction you would hope: the harder the paper, the lower the threshold tends to fall. When a specific question proves far more difficult than intended, that shows up in the data and it is taken into account.
External factors can matter too. In my experience, a compromised paper tends to push thresholds upward rather than downward, for reasons that become obvious once you think about what a leak does to the mark distribution. I have written separately about how Cambridge handles a leaked exam paper and what it means for candidates who sat it in good faith.
Two further things the Code of Practice commits to that are worth knowing. Statistical evidence used in grading relates to more than one previous series, not just last year. And the grading process must take account of comparability between alternative papers, including alternatives taken in different time zones — which is why variant thresholds differ without anyone being favoured. On that June 2025 Chemistry table, the A* threshold was 171 for option BX, 172 for BY and 173 for BZ. Same standard, three different raw marks.
Are boundaries decided before the exam?
No. Ofqual has stated that even in well-established qualifications, grade boundaries are never set in advance, because it is almost impossible to predict how much easier or harder candidates will find a paper — and even the examiners who write the papers find that challenging.
That same piece dismantles the mock-data prediction industry in one sentence. Organisations that publish "expected boundaries" from their own mock sittings do not know how representative their sample is of the national cohort. Multiply that problem by the size and spread of an international cohort and you can see why a poll of students who felt a paper was hard tells you about student experience, not about arithmetic.
My advice to students is blunt, and I give it every year. Do not overthink the grade boundaries, because there is nothing you can do about them. Focusing on your next step is worth far more than checking for news.
There is one twist here that I find encouraging. The students who chase boundary rumours hardest are usually the hardworking ones, and most of them do not look at boundaries until every exam is finished. When they come out of a paper that felt easy, they predict the threshold will rise — and then revise harder for the next one. The belief is wrong. The response to it is not.
Most of your grade boundaries were never judged by a human
Cambridge decides thresholds by evidence and judgement only at what it calls key grades. The thresholds for the remaining grades are calculated from those using arithmetic rules.
That is stated in the Code of Practice, and you can watch it happen in the published data. Take the June 2025 Chemistry table and check the six extended options against a single rule — B sits exactly halfway between A and C:
- BX: A 143, C 88 → midpoint 115.5 → B is 115
- BY: A 145, C 88 → midpoint 116.5 → B is 116
- BZ: A 146, C 88 → midpoint 117 → B is 117
- CX: A 145, C 88 → midpoint 116.5 → B is 116
- CY: A 146, C 89 → midpoint 117.5 → B is 117
- CZ: A 147, C 89 → midpoint 118 → B is 118
Six for six. Run the same test on D against C and E and it holds again — on Option CY, C is 89 and E is 67, and D lands on 78.
So if you are a B or D candidate, nobody sat in a room reviewing scripts at your threshold. It was interpolated. Ofqual describes the equivalent rule for UK GCSE grade 5 in the same blog post: divide the marks between grade 4 and grade 7 by three, adding any remainder to the lower grade width.
This is the fact that should end boundary speculation for good. A prediction about the B threshold is a prediction about a subtraction, and the only inputs are two numbers that will not exist until marking is finished.
Why there is no A* on Paper 4
Cambridge's threshold tables carry the line explicitly: grade A* does not exist at the level of an individual component. Look at the June 2025 Chemistry table and every component column stops at A. The A* appears only in the option rows, against the weighted total.
The same logic explains why your paper grades do not average into your syllabus grade. Cambridge calculates your syllabus grade directly from the total of the weighted marks on the components you took — not from the component grades. It goes further and says component grades are not used in determining candidates' results and are not part of them. Any paper-level letters you are shown are additional information, not your result.
Which means the intuition "I got A, A and B, so I must have an A overall" has no mechanism behind it. Two candidates with very different paper profiles and the same weighted total get the same grade. A candidate who is strong on Papers 2 and 6 and weak on Paper 4 can miss A* entirely, because in Chemistry, Paper 4 carries eighty raw marks against forty on each of the others.
One mark short — will they round you up?
No. A published threshold is a minimum qualifying mark, and if the A threshold is 146, then 146 gets an A and 145 does not.
But there is a real safety net, and almost nobody writes about it. Before results are issued, Cambridge runs a grade review: a targeted re-marking by senior examiners of the work of candidates most at risk of receiving inappropriate results. A candidate's work is re-marked at grade review if they are close to a syllabus-level threshold and at risk for at least one other reason. Cambridge also commits that where marks were adjusted downwards to offset a lenient examiner, and that adjustment leaves a candidate just below a threshold, the affected component is re-marked.
If you want your result reconsidered afterwards, the service is a review of marking rather than a re-mark, and the fee is waived if the enquiry changes your syllabus grade.
My practical advice differs from what most students expect to hear. Reviews and re-checks are not very successful most of the time, and they take time you may not have. In a small number of cases I have seen students gain two marks from a re-check — which tells you the realistic scale of the thing. Ask for one only if you are sitting close to the next grade boundary. Two marks changes nothing if you are twelve marks short. It changes everything if you are one mark short.
What if your 2026 exams were cancelled?
If you are in Kuwait, the UAE, Bahrain or Lebanon, your June 2026 sitting may not have gone ahead, and a different set of myths has grown up around what that means for grading.
The most common one deserves a direct answer: portfolio and contingency grades do not change the boundaries for candidates who sat written papers, and written-paper boundaries do not determine portfolio grades. The two groups are assessed separately against the same standard. If you sat a paper somewhere else in the world, your threshold reflects the performance of candidates on that paper.
If you are in an affected country, the thing that matters is knowing which route applies to you, because your revision strategy depends on it. I have set both out in detail — the portfolio of evidence route and how final grades are decided for Cambridge candidates, and the split between enhanced grading and international contingency grading for Pearson candidates. Read whichever applies before you plan another week of revision.
Does Edexcel IAL work the same way?
No — Edexcel International A Level is modular, reports UMS rather than PUM, and puts an extra condition on the A* that Cambridge does not.
Pearson explains the mechanics on its International A level results page. You get a UMS mark and a grade for each unit, then cash in for an overall grade. UMS is a conversion, not a percentage: Pearson's own illustration is that 53 raw marks in one session can represent the same achievement as 49 raw marks in the next, and UMS makes both count the same toward your final grade.
The A* rule is where students lose out. You need an A overall and at least 90% of the A2 UMS. For six-unit subjects such as Biology and Chemistry that means at least 480 UMS overall with at least 270 UMS from the A2 units. For four-unit subjects it is 320 overall with at least 180 at A2. Mathematics is different again: 480 overall, with at least 180 of the 200 combined marks from P3 and P4. A brilliant AS year cannot substitute for A2 performance — I have written a full breakdown of how to score an A* in Edexcel IAL exams if you are targeting one.
One resit detail worth checking carefully. Pearson states that for International A level it uses your best result from your two most recent attempts at a unit, provided you were entered under the same candidate identifier. In practice, students who retake usually improve, so this rarely bites. But if you are considering a third attempt at a unit you once did well in, do the arithmetic before you enter.
How to use past boundaries without fooling yourself
Past thresholds are a planning range, not a prediction. Used properly they are one of the most useful revision tools you have; used badly they produce false confidence.
Three rules I give my own students. First, grade a past paper with its own session's threshold, never with last year's — that is the only comparison that means anything. Second, look at several sessions to build a range rather than anchoring on one number, and target the top of that range plus a buffer. My IGCSE Biology 0610 boundary analysis shows why the safe target sits well above the average of recent sessions. Third, translate the total into per-paper targets, because a total tells you nothing about which paper is costing you the grade.
Then stop looking. Boundaries are the one variable in this entire process you cannot influence, and every hour spent refreshing them is an hour not spent on the papers that decide your raw mark. Working through solved past papers with full mark schemes will move your total. Reading boundary predictions will not.
If you want the underlying content secured first, the IGCSE Chemistry course runs live classes built around past-paper technique and mark-scheme language, and the IGCSE Biology and Chemistry notes eBooks cover both syllabuses in examiner wording. If you are unsure which route or target applies to you, message me on WhatsApp at +965 5137 5709 or email admin@chem-bio.info and I will point you to the right starting place.
Reviewed against the current Cambridge Code of Practice, the published Cambridge grade threshold tables and the latest examiner reports.
About the author
Hosni Showike is a Cambridge and Edexcel examiner and the author of five published science textbooks selling more than 20,000 copies a year across 40 countries. He has taught IGCSE and International A Level Chemistry and Biology for over fifteen years, and writes at chem-bio.info where he publishes grade threshold analysis, solved past papers and live exam-preparation classes. Every claim in this article is checked against the current Cambridge and Pearson documentation rather than second-hand summaries.
Frequently asked questions
Is 70% in IGCSE really an A*?
If the 70 you are looking at is a percentage uniform mark, it is a B, because the A* band starts at 90 PUM. If 70% is your raw mark, it may or may not be an A* depending on the session — the A* threshold in Chemistry 0620 has ranged from 74.5% to 86.5% of the raw marks across the last six sittings.
Do grade boundaries depend on how other students performed?
Cohort evidence is one of the inputs Cambridge uses, but not in the way students fear — thresholds are not set by ranking candidates or by fixing the share who receive each grade. Cambridge commits that its examinations are not norm referenced, so a classmate scoring highly cannot cost you a grade.
Why is my Cambridge AS grade an A when I have 90 PUM?
Cambridge International AS Level does not award an a*, so the percentage uniform mark range for grade a runs from 80 all the way to 100. A 90 PUM at AS is a very strong A, not a missed A*.
Should I request a review of marking if I am one mark below a boundary?
That is close to the only situation where it is worth doing, and the fee is refunded if your syllabus grade changes. Reviews succeed in a minority of cases and typically move a mark or two, so they are worth requesting when you are on the boundary and rarely worth it otherwise.
Try a free Class
