How do I revise for A-Level Maths?

A-Level Maths has a reputation as the exam you can’t bluff — and that is exactly why it is one of the most preparable exams you will ever sit. The content is fixed in a public national document, your board’s specification sets out the paper structure and mark schemes in advance, and the question styles return year after year in recognisable forms. Revising for A-Level Maths means, above all, using that predictability instead of rereading notes and hoping.

This guide sets out a four-step approach: know your exam, run an honest topic-by-topic diagnosis, turn your errors into marks, and finish with full papers under real conditions. It applies whether you sit AQA, OCR or Pearson Edexcel — where the boards differ, the specification is the referee.

01Know your exam before you revise for it

Every A-Level Maths course in England is built on the same national subject content, published by the DfE. Your exam board — AQA, OCR, Pearson Edexcel — turns that content into a specification: the document that fixes which papers you sit, how they are structured, what the assessment objectives are and how marks are awarded. Ofqual regulates the boards in England, alongside Qualifications Wales and CCEA elsewhere in the UK, and JCQ coordinates how assessment runs. For you, one document matters most: get your board’s specification and treat it as the contract for your exam.

The content itself divides into three strands. The pure core: proof, algebra and functions, coordinate geometry, sequences and series, trigonometry in radians, exponentials and logarithms, differentiation, integration, numerical methods such as Newton-Raphson and the trapezium rule, and vectors. Statistics: sampling, data presentation and interpretation — including work with the large data set — probability, the binomial and normal distributions, and hypothesis testing. Mechanics: quantities and units in the SI system, kinematics with the constant-acceleration equations and projectiles, forces and Newton’s laws, and moments. All three strands are examined; none of them is optional extra credit.

One more structural fact shapes everything: A-Level Maths questions are rarely single-topic. A mechanics question quietly demands calculus; a statistics question leans on algebra; a coordinate-geometry problem ends in a trigonometric equation. That is the modelling and problem-solving culture of the qualification — and it is why revising topics in isolation, from a book, in the same order every time, leaves you unprepared for papers that refuse to announce which chapter they are testing.

02Step 1: diagnose, topic by topic

The worst revision is doing what you already can: it feels productive and earns nothing. So before you plan a single session, take the specification checklist and ask one question per topic: could I do a standard exam question on this right now, unaided, without an example in front of me? Be strict with yourself — "I understand it when I see the solution" counts as shaky, not secure.

1. Respect the dependencies. Algebraic fluency carries everything: if rearranging, factorising and index laws are shaky, calculus and mechanics will keep collapsing under you. Repair foundations before summits.
2. Weight topics by exam value: a shaky core topic — differentiation, trigonometric equations, the normal distribution — comes before three peripheral ones.
3. Book a fixed weekly slot for the strand you quietly avoid. For many students that is vectors, proof or hypothesis testing; avoidance is a plan to lose those marks.

This diagnosis costs one afternoon and saves weeks of misdirected work. Remember the arithmetic of coverage: questions are spread across the whole specification, so twelve topics at a solid level beat five perfect topics and seven gaps — every time.

03Step 2: turn your errors into marks

A large share of lost marks comes not from questions that were too hard but from errors that were entirely known in advance — the same ones, every year, in thousands of scripts. The antidote is an error log. Every time a question goes wrong, don’t copy out the solution: record the error and its type — slip, misunderstood concept, misread question. Within two weeks you will see your personal error profile, and that profile is your revision priority list.

• Sign slips when rearranging: moving a term across the equals sign without flipping its sign, especially mid-way through a longer manipulation.
• Expanding (a + b)² as a² + b² — the missing middle term 2ab remains the most reliable mark-loser in algebra.
• Squaring both sides of an equation and never checking the candidates in the original — spurious solutions survive to the final line and cost the accuracy mark.
• Working in degrees where radians are required: the calculus of sin x and cos x only behaves in radians, and a calculator left in the wrong mode poisons every line that follows.
• Quotient-rule order: writing the numerator the wrong way round — it is vu′ subtracted from uv′ that ruins the derivative, and the error hides beautifully.
• Dropping the constant of integration in indefinite integrals — and then losing it again when solving a differential equation where the constant is the answer.

Then practise in short mixed sets: five or six parts spanning several strands, marked immediately, errors into the log. That is the regime of the real paper — a sequences question followed by a projectile with no warning — and it is what makes your methods reliable when the clock is running.

04Step 3: answer the command word, not the question you hoped for

Maths papers tell you exactly what they want — in the first word. "Calculate" means work out a numerical answer and show the relevant working. "Determine" asks you to establish a result from the information given. "Show that" hands you the destination and marks the journey: a structured argument or working that arrives at the stated result, with no steps skipped. "Derive" wants a sequence of logical steps building a result from known relationships. "State" is brief and precise, no argument needed; "explain" and "justify" are never answered with a bare number.

From this follows the single most profitable habit in mathematics exams: write your reasoning down, always. Name the method, show the substitution, keep intermediate results, and finish with a conclusion that answers the question asked. Mark schemes — published by your board alongside past papers — credit the method as well as the answer, and a legible line of working with one slip at the end routinely earns more than a bare correct number. In a "show that" question, the working is the answer.

05Step 4: full papers, under real conditions

In the final weeks, the balance shifts from learning topics to simulating the exam: complete past papers, in one sitting, timed, with only what the rules allow on the desk. Past papers with their mark schemes are the most realistic practice material in existence — no textbook exercise matches the mix, the phrasing and the tempo of the real thing. They are board-copyright, and the place to get them is your own board’s website, which also guarantees you are practising the right structure.

1. Sit your first paper early, as a diagnosis rather than a verdict: it shows how your knowledge behaves under time pressure, which no amount of note-reading measures.
2. Mark honestly against the mark scheme and count what the scheme counts: where did you lose method marks by skipping steps? Every skipped step goes into the error log.
3. Train your order: start where you are strong, bank those marks, keep a clock per question. Many candidates gain a grade’s worth of marks by changing nothing but their route through the paper.

06Statistics and mechanics are not optional extras

A quiet trap in Maths revision is spending ninety per cent of your time in the pure core because it feels like "real maths". The applied strands are examined with the same seriousness: hypothesis tests must be set up and concluded in context, the large data set rewards genuine familiarity, and mechanics questions want a clean diagram, consistent SI units and Newton’s laws applied like clockwork. Give both applied strands standing appointments in your week — they are dense with marks precisely because so many candidates under-revise them.

A-Level Maths does not reward talent; it rewards systematic preparation — a specification turned into a checklist, errors turned into a log, topics turned into timed papers. Take the structure of the qualification as your map and walk it at your own pace, strand by strand, week by week, until nothing on the paper looks unfamiliar. That is what a good grade looks like from the inside: no secrets, just coverage, method and margin.