17 August 2026 · Hugo Cheyne
A grade 3 sits at most 17 marks below a grade 4. An eleven-week plan for the November 2026 GCSE Maths resit, built on finding those marks.
Missing a grade 4 first time round is common, and it is recoverable. The first step is smaller than it looks: find out how many marks your child actually got.
On AQA's Higher tier in 2025, a grade 3 meant anything from 46 to 62 marks out of 240. A grade 4 needed 63. So a student holding a grade 3 is somewhere between one mark and seventeen marks short of a pass, and until the school tells you which, no revision plan can be written honestly.
This is the week-by-week plan for the resit itself. If the choice between a resit and a remark isn't settled yet, read resit or remark first. It assumes a November entry, and it's built around finding a specific, countable number of marks rather than re-teaching two years of content in eleven weeks, which can't be done.
Key Takeaways
- On AQA Higher in 2025, the gap between a grade 3 and a grade 4 was 17 marks out of 240. On Foundation it was 41. For an idea of how much work there is to be done, ask for the actual mark achieved, not just the grade. If there are only two or three marks in it, it may be worth requesting a remark as well.
- 26.1% of 17-year-olds in England passed GCSE Maths in the November 2025 series. Roughly one in four. The plan below is built around what separates them.
- The eleven weeks are not eleven equal weeks. Two blocks of holiday are where most of the work can be done, the term-time middle is thinner than families expect, and two administrative deadlines sit inside it.
- All three papers fall within six days, so everything has to be ready by 4 November (just after October half term).
- Resit results come back in January 2027, so your child will have started their post-16 course before the grade exists.
Before any revision happens, ask the school's exams officer for the mark, broken down by paper if they can. The grade alone can't tell you how big the job is.
Here is why, using AQA's published subject-level boundaries for GCSE Mathematics (8300) in the June 2025 series. Both tiers are marked out of 240.
| Tier | Grade 3 | Grade 4 | Gap |
|---|---|---|---|
| Higher | 46 | 63 | 17 marks |
| Foundation | 119 | 160 | 41 marks |
Source: AQA subject grade boundaries, June 2025 series, retrieved 17 August 2026.
Seventeen marks across eleven weeks is about a mark and a half a week. That's the whole task, stated honestly, and it's a very different proposition from "revise GCSE Maths again".
But the grade 3 band is wide. A student on 62 needs one mark. A student on 46 needs seventeen. Same grade on the same certificate, and a seventeenfold difference in the size of the job. Everything below depends on knowing which, which is why the first week is diagnosis rather than revision. Our guide to GCSE Maths grade boundaries has the full published tables and explains why boundaries move year to year.
One caution: these are AQA's 2025 figures. Boundaries are set after marking each year and the other boards set their own, so treat the gap as a close benchmark rather than a target line.
In the November 2025 series, 39,121 seventeen-year-olds in England sat GCSE Mathematics and 26.1% achieved a grade 4 or above, up slightly from 25.7% the previous November (JCQ, GCSE results - November 2025, Table 4, published 8 January 2026).
Roughly one in four. That is the base rate, and it's worth sitting with for a moment before building a plan, because it explains why the plan below spends its first fortnight on diagnosis and habits rather than on topics.
The same table goes on to explain that for eighteen-year-olds sitting the same papers in the same series, the pass rate was 19.3%, down from 21.3% a year earlier and well below the seventeen-year-olds. Those are students in Year 13, a further year from the last time anyone taught them the content, still attempting the qualification.
Roughly one in five. That's the case against letting it run into a second year.
The gap between those two figures is what carrying it costs. Pick a sitting, this November or next June, and work to it properly. What goes badly is the entry that rolls forward by default with no plan attached.
This is the standard advice, but the truth contains some nuance. The published boundaries show why it deserves an actual decision.
A student sitting Higher needs 17 more marks to reach a grade 4. The same move on Foundation takes 41. In pure arithmetic the Higher route asks for less than half as much.
But, as we know the two sets of marks aren't equally hard to get. Foundation questions have their difficulty capped, which is exactly why a grade 4 there took 67% of the marks in 2025 against 26% on Higher. Forty-one Foundation marks may well be more reachable than seventeen Higher ones for a student who found the Higher paper bewildering.
The honest framing is a question rather than a rule: which set of marks is more reachable in eleven weeks? For a student who was three marks off on Higher and understood most of what they saw, staying put is usually right. For a student who scored very little and spent the exam lost, Foundation is the better paper, and examiners say so directly. Eduqas reported Higher candidates in 2025 "gaining less than 15% of the available marks", where "entry to Foundation Tier would have probably led to greater success and more confidence gained" (Eduqas, GCSE Mathematics Examiners' Report, Summer 2025, retrieved 17 August 2026).
Two practical points. Foundation caps at grade 5, which is no obstacle when the target is a 4. And the entry is made by the school, so this is a conversation with the exams officer, not a decision to take at home. Have it in the first week, because it drastically changes what the eleven weeks contain.
For most students holding a grade 3, yes. The content is fresh, the gap is small, and everything below is built for exactly that case. But it's worth deciding rather than defaulting into, because schools often enter students as a matter of course.
The line to look at is the grade itself. A grade 3 on Higher means the gap is 17 marks at most. A grade 2 means it's more than 17 and possibly far more, on top of whatever went wrong the first time. Eleven weeks running alongside a brand-new post-16 timetable will not reliably close a gap that size, however good the plan is.
Entering anyway costs more than a wasted term. The result lands in January, in the middle of their first post-16 year, and a second fail is a great deal harder to come back from than a first. A student who has stopped believing the subject is possible for them won't revise for it in the spring, and at that point you've spent the confidence and bought nothing with it.
The June 2027 series is the same qualification with three times the runway (and the long Christmas and Easter holidays). A student continuing in full-time education has to keep studying GCSE Maths anyway under the 16 to 19 condition of funding, so choosing June isn't opting out of anything. It's using the whole year rather than the first eleven weeks of it.
The June route has a cost of its own. A summer entry means GCSE Maths competes with their new courses for a whole year rather than a single term, and a Year 12 spent with one eye on GCSE content is a weaker platform to build Year 13 on. So the closer they already are to a grade 4, the stronger the case for getting it finished in November.
Most resit plans divide the time into equal slices and hand one curriculum topic to each. That fails for two reasons: your child has already been taught all of it, and the calendar isn't flat.
From results day on 20 August to Paper 1 on 4 November is 76 days. Inside that sit two blocks of free time, a squeezed term-time middle, and two deadlines that have nothing to do with revision but will derail it if missed.
Two structural facts drive everything after this point.
All three papers fall inside six days. Wednesday 4, Friday 6 and Monday 9 November. There is no recovery time and only a chance to cram Paper 3 content after sitting Paper 2. The real deadline is the end of half term.
Results come back in January 2027. Your child will have started their post-16 course before the result exists.
| Week | Dates | The job |
|---|---|---|
| 0 | 20–23 Aug | Get the mark, per paper if the school can break it down. Size the gap against last year's boundary. Decide resit, remark or accept, and ask the exams officer to enter for November as insurance. |
| 1 | 24–30 Aug | Diagnosis, not revision. One past paper, cold, timed, marked strictly. Produce a list of lost marks sorted by how they were lost. Settle the tier question with school. |
| 2 | 31 Aug – 6 Sept | The four ways marks leak, before any topic work. Habits, not content. The cheapest marks on the paper are here. |
| 3 | 7–13 Sept | Term starts and time collapses. Set the weekly rhythm that has to last to November. Topic block one: the single biggest mark loser from week 1. |
| 4 | 14–20 Sept | Topic block two. First spaced revisit of week 2's habits and week 3's topic. |
| 5 | 21–27 Sept | Topic block three. Deadline: reviews of marking close 24 September. Confirm with the school in writing. |
| 6 | 28 Sept – 4 Oct | Deadline: November entries close in early October. Confirm the entry is actually submitted. Then a second timed paper: are the marks moving? |
| 7 | 5–11 Oct | Re-triage on the new evidence rather than the old plan. Are the habits now embedded? Start the next block. |
| 8 | 12–18 Oct | Mixed practice with the topic labels removed. This is where choosing the right method gets trained. |
| 9 | 19–25 Oct | A full paper in one sitting under exam conditions. Named gaps only from here. No new topics. |
| 10 | 26 Oct – 1 Nov | Half term. Three papers across the week, mirroring the real six-day cluster. Everything has to be ready by Sunday. |
| 11 | 2–9 Nov | Taper. Short retrieval only. No new content and no new past papers. Between papers, sleep and formula recall rather than post-mortems. |
Half term falls in week 10 for most schools and colleges, but not all. Check the college calendar before committing to that week, because it carries a disproportionate share of the plan.
For a student who needs seventeen marks, the fastest route is not new content. It's the marks already being lost on questions they can do. Examiner reports name the same four failures every year.
Not answering the question that was asked. Eduqas found "easy marks were lost through not checking the requirements of the question", with candidates asked to express 42 as a percentage of 300 instead calculating 42% of 300.
Working that can't be followed. Method marks exist, but only if the marker can see the method. Plenty of candidates showed extensive working with "no clear indication of what is being calculated".
No plausibility check. "Very few candidates employed checking strategies." The examiner's own test is one any parent can apply without knowing the maths: is £125 a reasonable price for a protractor?
Multi-step problems. "Multi-step problem solving questions will always be asked", and they're where grade 4 candidates most reliably come unstuck, because the student has to choose the method rather than execute a named one.
All quotations: Eduqas, GCSE Mathematics Examiners' Report, Summer 2025, retrieved 17 August 2026.
Spending week 2 on these four habits, before touching a topic list, is the highest return available in the whole eleven weeks. Our parents' guide to GCSE Maths covers what examiners say in more depth.
From week 3 the plan competes with a full post-16 timetable. A schedule that assumes two free hours a night will be abandoned by the end of September, so the unit is three sessions a week:
Two things about how that material is spread matter as much as the hours.
Revisit topics at increasing intervals rather than covering each once and moving on. Cramming into late October will feel more productive and leave less behind on 4 November.
And from week 8, take the topic labels off. Papers 2 and 3 don't announce which method a question needs, and recognising it is most of the difficulty. Practising mixed topics rather than one per session is uncomfortable, which is exactly why it works.
Every diagnostic above depends on one thing: a paper marked strictly against the mark scheme, with no marks awarded that weren't fully earned.
This is where the plan most often fails at home, and it isn't a discipline problem. Mark schemes award method marks for partial work, and deciding whether a half-finished solution earned one requires knowing what the examiner was looking for. A parent who doesn't know the curriculum can't make that call, and a generous mark in September produces a nasty surprise in January.
There are three ways round it. Ask the school whether a maths teacher will mark the week 6 and week 9 papers. Use the published mark schemes and be ruthless, accepting that borderline judgements will be wrong. Or have someone else do it.
That last one is what our GCSE Maths group sessions are for. They run to this structure, with the marking and the re-triage built in. If you're weighing up whether the gap is one you can close at home, the signs a child needs a tutor sets out what to look for.
On AQA Higher tier in 2025, a grade 3 needed 46 marks out of 240 and a grade 4 needed 63, so the gap was at most 17 marks. On Foundation the same gap was 41 marks, from 119 to 160. Because the grade 3 band is wide, a student at the top of it may need only one mark, which is why the actual mark matters more than the grade.
In November 2025, 26.1% of 17-year-olds in England achieved a grade 4 or above in GCSE Mathematics, from 39,121 entries, against 25.7% the previous year. The rate for 18-year-olds was 19.3%. Figures are from JCQ's Table 4 for the November 2025 series, published 8 January 2026.
It depends on how they found the Higher paper rather than on a general rule. The gap from a grade 3 to a grade 4 is 17 marks on Higher and 41 on Foundation, but Foundation questions are less demanding, so those 41 marks may be more reachable. A student who scored very little on Higher is usually better served by Foundation. The entry is made by the school, so discuss it with the exams officer early.
It's enough to move a grade, and it isn't enough to re-teach two years of content. Which is fine, because the task is finding at most 17 marks rather than relearning the course. The students who manage it generally close named gaps identified by a marked paper, rather than revising in general.
November suits a student who was inside the grade 3 band, where the gap is 17 marks at most on Higher and the content is still fresh. If they scored below a grade 3, the gap is wider than what eleven weeks will reliably close alongside a new post-16 timetable. Also a second fail in January is harder to recover from than the first. June 2027 gives roughly three times the runway. They have to keep studying GCSE Maths either way under the 16 to 19 condition of funding, so waiting for June isn't opting out.
Three sessions a week: two short retrieval sessions and one longer mixed practice session. It's a realistic target that survives a full-time post-16 timetable. Consistency across eleven weeks matters more than volume in any single week, and a plan that assumes more than this is usually abandoned by late September.
Nothing new. The papers fall on 4, 6 and 9 November, and there's no useful way to revise Paper 3 content in the 48 hours after Paper 2. Short recall of formulae, sleep, and no post-mortems of the paper just sat. Everything that will affect the result was done by the end of half term.
Sit in on a lesson led by one of our expert PGCE qualified teachers.