Anciennes Épreuves: How to Find Them and How to Use Them

Most candidates collect anciennes épreuves and then use them almost entirely wrong.

They save them for the final fortnight, treat them as a test of whether they are ready, score badly, and feel worse. The papers were the single most useful thing they owned and they got about a tenth of the value out of them.

Past papers are not an examination you sit at the end. They are the syllabus, the marking guide and the study plan, and they are most useful at the beginning. Here is where to find real ones and how to work through them properly.

Where to actually find them

In rough order of how much you should trust what you get:

Source	Worth knowing
The institution itself	Some schools sell or publish their own past papers. Always check here first, because these are definitive.
Your school or department	Teachers and departmental libraries often hold collections that never reach the internet.
Recent candidates	People who sat it last year remember what came up and often kept their papers. Underused and free.
Bookshops near campuses	Compiled volumes are common around major institutions.
Online repositories	Wide coverage, uneven accuracy. Useful for volume, verify before trusting.
Messaging groups	Fast and frequently wrong. Treat anything from here as unverified.

The eStudyUniverse app exists partly for this reason: papers downloaded once and readable afterwards without data, because the moment you most want to revise is rarely the moment you have bundle left.

Check that what you downloaded is real

Circulated papers are sometimes mislabelled, sometimes retyped with errors introduced, and occasionally invented. Revising from a wrong paper is worse than not revising, because you learn a pattern that does not exist.

Four quick checks before you commit time to a paper:

• Does it name the institution, the year and the series? A paper with no identifying header is a warning sign on its own.
• Does the structure match other years? If every other paper has three sections and this one has five, something is wrong with this one.
• Is the register right? Official papers have a consistent, slightly stiff house style. Retyped copies often read casually and contain typing errors that the original would not.
• Do two independent sources have the same paper? If yes, you can rely on it.

Be especially careful with answers. Papers circulate widely with unofficial solutions attached, written by whoever compiled them, and those solutions can be confidently wrong. A question from an unofficial compilation is usually safe. An answer from one is not.

The order that gets the value out

First, read one without attempting it

Before you try to answer anything, read a whole paper through as a document. You are looking at shape rather than content: how many sections, how the marks are distributed, how much writing each answer seems to want, what the instructions actually permit.

Ten minutes here removes a surprising amount of exam day confusion, and candidates skip it because it does not feel like work.

Second, attempt one badly, early

Sit a paper months before you feel ready. You will do poorly. That is the entire point, and it is why doing this early matters.

A bad score in month one is information you still have time to act on. The same score in the final fortnight is just distressing. Waiting until you feel ready means waiting until the information is nearly useless.

Third, build the frequency tally

This is the step that does the most work and almost nobody does it.

Take every paper you have. For each question, write down its topic. Then count. You will end up with something like this:

Topic	Appearances in 6 years	What that tells you
Topic A	6	Certain. Must be strong here.
Topic B	5	Near certain. Same.
Topic C	3	Regular. Worth solid coverage.
Topic D	1	Occasional. Cover, do not perfect.
Topic E	0	Check it is still on the syllabus at all.

That table is your revision order, and you derived it from evidence rather than from a feeling about what seems important. It takes an afternoon and it directs months.

One caution. A topic absent for several years may have been removed from the syllabus, or may be overdue. The tally tells you where to look, not what to conclude, so check the current syllabus before deciding a topic is dead.

Fourth, work by topic, not by paper

Now stop sitting whole papers for a while. Instead, pull every question on one topic from every year you have and do them together.

Doing six years of questions on the same topic in one sitting shows you the pattern immediately: the same idea asked four slightly different ways, the same trap set twice, the phrasing the examiner keeps returning to. You cannot see any of that doing one paper at a time.

Fifth, return to full timed papers near the end

Save two or three complete papers, untouched, for the final stretch. Sit them in one go, timed, without notes, ideally at the time of day you will actually sit the real one.

These are not for learning content any more. They are for rehearsing pacing and concentration across the full duration, which is a separate skill and the one most likely to fail you.

Marking yourself honestly

Everything above collapses if you mark generously, and generous marking is the default.

Two rules. Half remembered is wrong. If you needed a hint, could not have written it in the hall, or arrived at the right answer by a route you could not repeat, mark it wrong. Right answer, wrong method is wrong. It will not survive the next question that changes one detail.

Then do the part that actually converts practice into marks. Every error becomes a specific item you review on a schedule, not a vague intention to be more careful. Our guide to spaced repetition covers how to run that review so the same mistake stops returning.

Keep a running list of repeated errors, in your own words. Read it before every timed attempt. Most candidates make the same five mistakes for months and never once write them down.

Four ways people waste past papers

1. Reading the question, then the answer. This feels efficient and teaches almost nothing. The struggle before the answer is where the learning happens.
2. Saving them all for the end. Then there is no time to act on what they reveal.
3. Doing them untimed, always. You build knowledge and never build pace, then discover the gap on the day.
4. Collecting instead of doing. Twenty downloaded papers and three attempted is a very common ratio and a completely useless one.

Working from a phone, on limited data

Most advice about past papers quietly assumes a printer, a desk and a connection. If you have none of those, the method still works, but a few things change.

Download once, work offline. Get the papers onto the device while you have connection and open them later. The times you most want to revise, in the evening, on a journey, during a power cut, are exactly the times you are least likely to have bundle to spare.

Write your answers on paper anyway. Reading a question on a screen and thinking through the answer is not the same exercise as writing it out. The exam asks you to produce a written answer under time, so practise producing a written answer under time. A cheap exercise book is enough.

Keep the tally in a notebook, not an app. It is faster to add a line to a page than to open something, and you will look at it more often. The frequency tally is only useful if you actually consult it.

Batch what needs connection. Verifying a paper, checking a syllabus, downloading a new year. Do those together in one session rather than one at a time, and spend the rest of the week working from what you already hold.

After the exam, keep them

Two reasons, and both get forgotten in the relief of having finished.

If you sit it again, your own attempts are the most useful revision material you will ever own. They show precisely where your marks went, which is information no general guide can give you. Candidates who retake usually start over from scratch when they are holding a detailed diagnosis of their own weaknesses.

And write down what came up while you still remember it, ideally the same day. Your recollection of this year’s paper is genuinely valuable to the candidates coming behind you, including your own siblings and juniors. Almost nobody does this, which is why the most recent year is usually the hardest one to find.

Questions candidates ask

How many years should I get?

Five or six is plenty for a frequency tally to mean something. Ten is not twice as useful, and very old papers can reflect a syllabus that has since changed. Prefer recent years over volume.

What if there are no official answers?

Common, and not fatal. Work the question, then verify against your textbook or a teacher rather than against a circulated solution sheet. For factual questions your notes settle it. For extended answers, showing your attempt to someone who teaches the subject is worth more than any answer key.

Do papers from a different school help?

For subject practice, yes. For learning a particular institution’s habits, no, and that second thing is most of the value. Use other schools’ papers to drill a weak topic, and your target school’s papers to learn what it actually asks.

Is memorising past answers worth anything?

Very little, and it is a trap. Papers repeat topics, not questions. A memorised answer to a question phrased slightly differently is often worse than no preparation, because you will answer the question you learned instead of the one in front of you.

Start with what you already have

You do not need a full set to begin. Take whatever papers you can reach today, read one straight through without answering it, and start the tally with those.

Add more as you find them. The tally gets sharper with every paper, and you will be revising in the right order from the first week rather than the last.

More on how ranked entry changes what you should prioritise in how a concours actually works, and more preparation guides in Exams and Past Papers.