How to use GCSE Maths past papers properly — most students get this wrong

There's a comforting myth that "doing past papers" is what revision looks like in the final stretch of Year 11. Sit down, do paper, mark it, do another, repeat. It feels productive — there's a stack of completed papers, the numbers go up most weeks, and the student is visibly working. But a worrying number of students who do this hit a ceiling around the same grade they started at, sometimes lower. Volume of papers is not the same as learning from papers.

This post is about the difference. The students who go up a full grade in the final months of Year 11 are usually doing fewer papers than the ceiling-stuck ones — but they're using each paper much, much better.

The trap: doing papers without learning from them

The pattern that doesn't work, in detail:

1. Monday: do a Higher Paper 1, get 80/240.
2. Tuesday: mark it quickly, look at the score, feel slightly disappointed.
3. Wednesday: do a Higher Paper 2, get 78/240.
4. Thursday: mark that one too.
5. Friday: do Paper 3.
6. Repeat for the next four weeks.

The student gets through 10 papers and the average score doesn't really move. The marks they're losing are mostly the same marks every week, on the same kinds of questions — but because they're not pausing to do anything about it, the mistakes calcify rather than correct. Five weeks later, the student is exactly as good at the things they were already good at, and exactly as bad at the things they were bad at.

The mistake isn't doing past papers. The mistake is treating each paper as a one-shot performance, instead of as the start of a feedback loop.

The loop that actually works

Here's the alternative — what we'd call the four-step paper loop:

1. Do the paper under timed conditions, marking your own working as you go (more on this below).
2. Mark with the mark scheme. Not against an answer key — against the actual AQA mark scheme, line by line.
3. *Identify the topics, not the questions, that lost marks.* A question on compound percentage change becomes "percentages — compound growth — reverse calculation". That's the entry on your list, not the question number.
4. Drill those topics, not those questions, before doing the next paper. Practice 10 questions on compound percentages from a topic resource. Then a few days later, do the next paper.

This loop is how a student who's done 3 papers but used them well outperforms a student who's done 10 but used them as performance reps. The first student's weak topics are closing; the second student's are not.

A practical anchor: in weeks 1–4 of any final revision push, do maybe 1 paper per week, with 4–5 days of topic drilling in between. Only in the final 2 weeks should you ramp up to 2 papers a week, and only because by then you've closed most of the topic gaps.

Why doing the same paper twice is mostly useless

A surprising number of revision plans recommend re-doing a paper a few days later "to see if you've improved". It feels reasonable. It mostly isn't.

The problem is that 24–72 hours after first marking a paper, students remember the answers to many of the questions. Not the methods — the answers. Re-doing the paper measures memory of specific answers, not improvement in underlying skill. The second-attempt score is almost always higher; the actual capability is mostly unchanged.

If you want to test whether a topic has actually improved, do questions from a different paper on the same topic. The questions will look different — different numbers, different wording, different contexts — and the student either has the skill or doesn't. That's the real test.

The one exception: re-doing a paper after 3+ months can be useful if you'd genuinely forgotten the questions, and only as a stamina check, not as a learning tool.

Timed vs untimed — when each is right

This is the question students ask most. The honest answer is that both have their place, at different points in the revision arc.

Untimed papers are the right tool early in revision. The student is allowed to take as long as they need, look up formulas, even pause and think. The goal isn't speed — it's learning what the paper actually tests, what types of questions appear, and what techniques you need. A first untimed pass through a recent paper is sometimes more like a guided tour than a test.

Timed papers are the right tool in the final 2–3 weeks. By this point the underlying skills should be in place; what you're training is stamina, pacing, and decision-making under pressure. Skip a hard question and come back to it — that's a skill. Recognise a question type quickly from the wording — that's a skill. These only get trained when the clock is running.

A common mistake is to do timed papers from the start of revision. Students who do this often build up associations between "past paper" and "stress" before they've built up the underlying competence. The numbers stay low, confidence erodes, the whole exercise turns sour. Untimed first; timed later.

The role of the mark scheme

If we had to name one habit that separates students who improve from students who plateau, it would be this: read the mark scheme in detail, line by line, after every paper.

The AQA Maths mark scheme is not a list of correct answers. It's a description of how marks are awarded — which working steps earn which marks, what counts as an acceptable alternative method, what specific phrasing is required for "reasoning" marks. Most students glance at the answers and move on. The students who actually move grades read the mark scheme and ask: "did I get the same answer for the same reason?"

A worked example will land this better than abstract description. Consider a quadratic question worth 5 marks:

• M1 — for setting up the equation correctly
• M1 — for substituting into the quadratic formula correctly
• A1 — for correct discriminant value
• A1 — for one correct root
• A1 — for both correct roots

A student who only writes down the final two roots and gets one wrong scores 1/5 — they get one A1 for the correct root, nothing else. The same student who shows working — equation, formula, discriminant, then both roots, with one wrong — scores 4/5. Same final answer. Different mark.

This is the method mark vs accuracy mark distinction. M1 marks are awarded for showing a correct method, regardless of whether the final answer is right. A1 marks are awarded for accuracy. B1 marks are independent — usually awarded for stating a result or fact without working. A student who knows this and writes down every line of working is claiming method marks they couldn't otherwise get — sometimes 8–12 extra marks across a paper, which is a full grade boundary's worth.

The other reason to read the mark scheme: it tells you what counts as a valid reason on geometric reasoning questions. "Alternate angles" is valid. "Z-angles" isn't. "Co-interior angles add to 180°" is valid. "C-angles" isn't. Students lose 2-mark reasoning marks all the time because they wrote a folk version of the rule rather than the formal one. The mark scheme is the authoritative reference.

Common student leaks where they shouldn't lose marks

A few specific patterns we see repeatedly when working through students' marked papers:

• Not writing down the formula before substituting. Costs B1 marks on questions where "stating the formula" is independently awarded.
• Rounding too early in multi-step calculations. The mark scheme often accepts a range, but the range is narrow — round to 3 significant figures only on the final answer, never intermediate steps.
• Skipping the units on the final answer. "12.5" is a wrong answer if the question asked for cm². "12.5 cm²" is right.
• Not showing the method when the question says "show that". "Show that the value is 25" requires the working that leads to 25, not just writing 25. Students who shortcut this lose all the method marks.
• Misreading "leave your answer in surd form" or "give your answer to 3 significant figures". These are explicit instructions; a numerical answer in the wrong form is sometimes a full A1 loss even when the underlying maths is correct.

A useful habit during the paper loop step: after marking, write down two or three of these patterns the student fell into on this paper. Over time the list shrinks as the student internalises them.

Where to get past papers

You have several routes:

• AQA's own site publishes recent question papers, mark schemes, and grade boundaries for free. Our GCSE past papers hub indexes the main exam boards (AQA, Edexcel, OCR, WJEC, Eduqas, CCEA) and links straight to each board's official PDF page. This is genuinely sufficient material for most students.
• Older AQA papers (pre-2017, in the old A*-G system) are available but less useful — the specification changed in 2017 and the question style is different. Stick to 2018-onwards for most practice.
• ExamVault's [Practice Pack](/gcse/packages) at £40 ships 10 brand-new ExamVault-authored papers with built-in mark schemes and adaptive feedback. The advantage over the free official ones is that each question is tagged to a topic, so the "drill the weak topics" step in the loop above happens automatically — you can jump from a wrong answer straight to 10 more questions on the same topic. If cost is a factor, the free official papers are absolutely adequate; the Practice Pack is the convenience version.
• ExamVault's [Mock Pack at £20](/gcse/packages) is a different product: 5 strict timed mocks designed for the final 2-week stretch, with no "stop and review" functionality. It's the exam-day simulator, not a learning tool.

How this connects to the wider revision plan

If you're putting together a final-stretch revision plan, the paper loop is one part of a wider rhythm. We have a separate post on the 6-week GCSE Maths plan that walks through how to slot past papers into a structured week-by-week build-up. The short version: do the diagnostic paper in week 1, drill topics in weeks 2–4, do mock papers in week 5, and use the final week for maintenance only.

The Adaptive GCSE Maths subscription at £30/month — £1 a day — is designed to replicate the "drill the weak topics" step automatically. Each day's pack pulls 10 questions weighted toward the topics where the student's mastery score is lowest. It doesn't replace past papers (you still need timed full-paper practice), but it makes the "drill between papers" step efficient.

The honest bottom line

Past papers are the most valuable revision resource in the system, and the most commonly wasted. The students who turn them into grade movement do four things differently:

1. They do fewer papers, with proper review between each.
2. They mark with the mark scheme, line by line, not against a numerical answer key.
3. They identify the topics losing marks, then drill those topics from other resources before the next paper.
4. They learn the method-mark vs accuracy-mark distinction and write working accordingly.

Do all four and a student who'd otherwise plateau at grade 5 routinely lifts to grade 6 in the final two months. Skip them and the stack of completed papers stays impressive, while the actual grade refuses to move.