Navigating Volatile Grade Boundaries
Why aiming for a standard specification threshold is a high-risk strategy for A-Level candidates.
Every summer, a familiar conversation happens on results day.
A student worked out they needed 60% for a grade. They scored 62%. And they still missed it.
They did nothing wrong. They just planned around a number that was never fixed in the first place.
Grade boundaries are not set in advance
This is the part most students never get told properly.
The boundary for a grade is not decided when the specification is written. It is not decided when the paper is written. It is not even decided when you sit the exam.
It is decided after every script in the country has been marked.
The exam board looks at how the whole cohort performed. If the paper turned out to be hard, the boundaries come down. If it was gentle, they go up.
So when you sit down in the exam hall, nobody in the world knows what a grade A is worth yet.
Why the boundaries move
The system is called comparable outcomes.
The idea is fairness across years. A student who performs as well as last year’s grade 7 candidate should also get a grade 7, even if this year’s paper was much harder.
The only way to do that is to move the number.
It sounds sensible, and it mostly is. But it has one consequence that matters enormously for how you revise:
The target you are aiming at is invisible until it is too late to hit it.
What this looks like in real numbers
Let’s use Edexcel A-Level Maths (9MA0). It is marked out of 300 across three papers.
Look at what the A* boundary has done:
| Series | A* boundary | As a percentage |
|---|---|---|
| June 2022 | 217 / 300 | 72% |
| June 2023 | 244 / 300 | 81% |
| June 2025 | 258 / 300 | 86% |
| June 2026 | 254 / 300 | 85% |
That is a swing of 41 marks for the same grade, on the same qualification, in four years.
A student who trained to hit 60% would have comfortably earned an A in 2019, when the A boundary was 165. The same performance in 2025 would have been a B, because by then an A took 214.
Two series are missing from that table on purpose. In June 2019 and October 2020, Pearson published no A* component boundary at all. The grade was awarded; the figure was simply not published.
If you want the full picture, with every series set against the share of students who actually got each grade, it has its own article: why A-Level maths grade boundaries keep going up. The short version is that the marks needed moved 45 between 2019 and this summer, and the share getting an A moved nine tenths of a point, which tells you what the boundary is really for.
GCSE is no calmer
Edexcel GCSE Maths Higher (1MA1) is marked out of 240.
Watch the grade 7 boundary over three consecutive summers:
- 2024: 137 marks
- 2025: 156 marks
- 2026: 146 marks
It jumped 19 marks, then dropped 10.
A student in 2025 who planned around the 2024 number would have been 19 marks short, roughly a full exam question and a half, while believing they were safely over the line.
Why it jumped is worth knowing, because the obvious answer is wrong. It was not a stronger year group: exactly the same share of students got a grade 7 in 2025 as in 2024. Why are grade boundaries going up? sets the marks needed against the share who reached them, series by series.
The grade 9 boundary moves even more, because it is designed to pick out the top few percent nationally rather than a fixed score.
Every GCSE series back to 2017 is on our GCSE maths past papers page, split by paper and by tier.
The mistake almost everyone makes
Here it is, plainly:
Using last year’s boundary as this year’s target.
It feels rational. It is the only hard number available. Every revision website publishes it.
But last year’s boundary describes last year’s paper and last year’s cohort. It has no authority over yours.
Treating it as a target is like packing for a trip using last week’s weather forecast.
Build a buffer instead
The fix is not complicated. It is just uncomfortable.
Take the highest boundary from the last five years. Then add a cushion on top.
For Edexcel A-Level Maths, if you want an A, do not train to 71%. Across the six normal sittings since 2019 the A boundary has run from 164 to 214 out of 300, so 214 is the number to beat, not the 165 that would have done it in 2019.
For GCSE Maths Higher grade 7, do not aim at 146. Plan for around 155 to 165 out of 240.
Yes, this means working harder than the minimum. That is the point.
A buffer is what turns a grade from a gamble into an expectation.
Know which boundary you are actually reading
A quick but important technical note, because this catches people out constantly.
Exam boards publish two sets of numbers.
Subject boundaries apply to your total across all papers. This is the one that decides your grade.
Notional component boundaries are published per paper. They exist for internal reporting and school analysis.
For Edexcel A-Level Maths in June 2026, the notional Paper 1 boundaries were 81 for an A*, 67 for an A and 55 for a B, out of 100.
Your grade does not come from these. A weak Paper 1 can be rescued by a strong Paper 3, because only the total counts.
Students often panic after one bad paper. They shouldn’t. There are two more still to bank.
Stop leaking marks you have already earned
Boundaries are outside your control. Marks are not.
In my experience marking student scripts, the marks lost to genuine gaps in knowledge are usually fewer than the marks lost to these:
Answering a slightly different question. The paper says normal, the student differentiates and uses the tangent gradient. The paper says rate of change, the student evaluates the function instead of differentiating it.
Abandoning the final part. Later parts of a question often carry method marks that do not depend on getting the earlier parts right. Walking away from them gives up marks that were free.
Ignoring the information in the question. If a question states that an angle lies between 0 and π/2, and your answer comes out negative, the question has just told you that you are wrong. Use it.
Assuming one paper is the soft one. Across the normal June sittings the A* boundary on a single paper has ranged from 70 to 89 out of 100. June 2024 wanted 89 on Paper 3, the Statistics and Mechanics paper, and 81 on Paper 1. Whichever paper you have decided is the easy one, some series disagrees with you.
Not checking the given range or form. “Give your answer in the form ax + by + c = 0 where a, b and c are integers” is an instruction, not a suggestion.
Ten marks recovered from habits like these will outrun any boundary movement you are likely to see.
If you land just below a boundary
You can ask for a Review of Marking.
Three things to understand before you do.
First, it is not a re-mark from scratch. A senior examiner checks whether the mark scheme was applied correctly.
Second, marks can go down as well as up. The whole script is reviewed, not just the parts you query.
Third, it costs money and takes time, and time matters if a university place depends on it.
The honest advice: request it when you are one or two marks below a boundary and you can point to a specific question where the method looks sound. Do not request it as a general hope.
Ask your teacher to look over the script first. A good subject teacher can usually tell within twenty minutes whether there is a real case.
The short version
Grade boundaries move, sometimes by a lot, and you cannot know them in advance.
So stop aiming at them.
Aim above the worst case, protect the marks you already know how to earn, and let the boundary land wherever it lands.
That way, results day is a confirmation rather than a coin flip.
Boundary figures quoted are from Pearson Edexcel’s published grade boundary documents. Always check the official Pearson website for the exact qualification code, 9MA0 for A-Level Maths and 1MA1 for GCSE Maths, rather than relying on summaries.