How Many Past Papers Should You Do Before GCSE Maths?

What one past paper actually costs

Ninety minutes to sit it. Thirty to forty-five to mark it properly against the mark scheme. Forty-five to sixty to rework what went wrong. Call it three hours per paper, which makes twelve papers about thirty-six hours — a full term of maths revision on its own, and the reason the answer is twelve rather than twenty-five.

This is also why students who report having done twenty papers have almost always sat twenty and processed six. Sitting is the enjoyable part: it takes a defined ninety minutes, it produces a score, and it feels like an achievement. Marking and reworking are unpleasant, undefined, and where the entire benefit sits.

If you have three hours and a choice between sitting a new paper and properly finishing the last one, finish the last one. Every time.

The schedule

The count is less useful than the distribution. Twelve papers in the last month is a measurement exercise; twelve spread across four months is a training programme.

The final two weeks surprise people. Sitting papers has by then stopped teaching anything new, and the return per hour on reworking old mistakes is several times higher.

• Before February — no full papers. Exam questions by topic, in sets of six to ten, with the mark scheme afterwards.
• February to Easter — one full paper a fortnight, marked within a day of sitting it.
• Easter holiday — two or three papers, and sit one complete set of three on consecutive mornings to rehearse the shape of the real fortnight.
• Easter to two weeks out — one paper a week, timed, at the time of day the real paper starts.
• Final two weeks — none. Rework the wrong answers from the previous eight papers instead.
• Between Paper 1 and Paper 2 — two or three weeks usually separate them, and this is working time, not a break.

Marking is the lesson

Mark against the scheme rather than against the answer line. GCSE mark schemes award method marks, accuracy marks and independent marks separately, and a student who compares only final answers will record a question as wrong when they lost one mark of four, or as right when they were fortunate. Where you cannot see why a mark was lost, that is the question to take to a teacher — it is a far better use of a lesson than 'can you go over percentages'.

Then the rework rule: cover the scheme, take a blank page, and do the question again from the start, on a different day. Reading a worked solution and nodding is the commonest way to waste a past paper, and it feels like revision because the solution makes sense as you read it. Making sense of someone else's solution and producing your own are different acts, and only one of them is examined.

The list that comes out of it

Keep a running tally of lost marks by topic. After four papers it stops being random: six to ten topics will be producing most of the losses, and they are usually not the topics the student says they find hard.

That list is the actual output of the whole exercise. The papers are diagnostics; the list is the treatment plan, and the remaining weeks belong to it. A student who sits twelve papers and never writes the list has bought twelve measurements of a problem and no attempt to fix it, which is why their scores plateau after paper five and stay there.

In our GCSE groups the list is written on the same sheet every week and topics are crossed off only when a question of that type is answered correctly, cold, a fortnight later. Crossing something off after one successful attempt is how a topic reappears in June.