How to Study for an Exam: A Research-Backed Plan

Spreading study across several days usually beats packing it into the last one, and it usually feels like the opposite. In three flashcard experiments published in 2009, in which people learned word pairs of the kind tested on a US graduate admissions exam, the psychologist Nate Kornell found that spacing practice out worked better than massing it for 90 percent of participants, yet after the first session 72 percent believed massing had worked better.¹

So how to study for an exam starts with the calendar. Count back from the exam date and give each of five steps its own stretch of time:

1. Map what the exam covers
2. Spread your sessions across the weeks you have
3. Test yourself with practice questions from the start
4. Mix topic types once you can do each one
5. Use the last days for review, not new material

Prefer to watch? The video covers the same ground as this article.

VideoHow to Study for an Exam: A 5‑Step Plan That WorksWiserHours

How to study for an exam: start from the date

A study plan should start from the exam date because the right moment to return to a topic depends on how long you need to remember it. Nicholas Cepeda and his co-authors drew exactly that planning rule from their 2008 spacing experiment: before you divide up your study time, decide how long the material has to last.²

The plan in this guide is meant for any timed written exam, whether that is a university final, an A-level, a bar or accountancy exam, or a national entrance test. It is built on the two techniques that John Dunlosky’s team, comparing 10 study methods in a 2013 review, rated most useful: testing yourself, and study spread over time, which researchers call distributed practicedistributed practice: Spreading study or practice of the same material across several sessions with gaps between them, rather than cramming it into one. Also called spacing, it tends to give longer-lasting learning than the same time spent in a single block.Full entry in the glossary.³ Why those two work is covered in how to learn more efficiently, according to the research; the job here is to put them in order on a calendar.

Without a plan, most study drifts toward the deadline. Dunlosky’s review describes a pattern some researchers call the procrastination scallop: in a 1971 study, students who were tested only every three weeks studied more and more as each test approached, while daily tests kept their study steady.³ Think of a student with finals in six weeks who means to start “properly” next Monday, then the Monday after. By the time the exam feels close enough to act, there is no room left for gaps between sessions.

The study

Moderate evidence

One review session, final tests up to a year later

Cepeda’s team tested 1,354 volunteers from an online research panel, aged 18 to 72 and living in many countries. Each one studied 32 little-known facts, came back for a single review after a delay of up to about three and a half months, then sat a final test as much as a year after that review. The ideal delay lengthened as the final test moved further off, yet shrank relative to the total wait: roughly 20 to 40 percent of the wait when the test was a week away, 5 to 10 percent when it was a year away. Reviewing too soon cost far more than reviewing too late.²

Trivia facts and one review session are simpler than a syllabus, so read the percentages as a sense of direction, not a formula. The authors also warn that squeezing a topic into too short a period gives misleadingly high mastery at first that does not last.² In practice, that means two things. Give every topic its first pass early enough to leave a real gap before you come back to it, and when the exam is still weeks away and you are torn between coming back in a few days or in a week, choose the week.

The five steps, from first week to exam day

The five steps run in a fixed order because each one feeds the next: the map tells you what to schedule, the schedule makes room for self-testing, and self-testing shows when a topic is ready to be mixed with others. The last days then go to review.

1. FirstMap the exam and rate each topic
2. Early weeksFirst pass of every topic, each opened with questions
3. Middle weeksSpaced returns; mixed practice once topics are familiar
4. Final weeksTimed past papers across all topics
5. Last daysReview what you know; no new material

The order matters more than the exact dates, so a plan with less time compresses each stretch rather than dropping one.

1. Map what the exam covers

Start by finding out exactly what the exam will test, since you may not need to revise everything, according to the University of Auckland’s study guide. It suggests checking the course outline, asking your lecturer or tutor, and reading past papers to see the format, such as multiple choice or essays.⁴ The University of Bath adds that the weightings of questions are worth knowing, and that your teacher has the up-to-date details.⁵ For a professional or national exam, the equivalent is the official syllabus, specification or content outline, whatever your exam body calls it.

Then decide how to share out your time. Singapore Management University’s guide suggests giving fewer sessions to subjects you are confident in and more to weaker ones.⁶ The catch is that confidence is a poor gauge. In the second of Henry Roediger and Jeffrey Karpicke’s 2006 experiments, students who reread a passage felt more confident about remembering it than students who tested themselves, yet remembered less a week later.⁷

A candidate for an accountancy exam, say, might list every syllabus area, mark the ones that carry the most marks, and then answer three or four quick questions on each before rating it. Rate topics by what you can answer.

2. Spread your sessions across the weeks you have

The same hours spread over separate days do more than those hours spent in one block. A meta-analysismeta-analysis: A study that combines the results of earlier studies on the same question into one overall estimate. Pooling makes the estimate more precise, but it cannot repair the studies it pools: a meta-analysis of surveys is still survey evidence.Full entry in the glossary that Cepeda’s group published in 2006, covering 317 experiments, found an advantage for spaced over massed study whether the final test came within a minute or more than a month later, and found that the ideal spacing widened as the test moved further away.⁸

On our reading, that gives each topic at least three visits on different days: a first pass, a return after a gap, and a later review. For an exam a few weeks away, the returns come after days rather than hours; for one several months away, after weeks.² In the 2006 review, tests six months or more after the last session were best served by gaps of at least a month.⁸

That schedule is an untested illustration: stretch or squeeze the blocks to fit your own exam date and number of topics.

1. Too short: reviewing straight after the first session did worse than the best gap, and lengthening the gap helps steeply at first
2. Best gap: the peak: for an exam soon, it comes after a fairly short gap
3. Too long: past the peak, memory falls off gently, so erring long costs less than erring short
4. Exam further away: the peak moves to a longer gap, so the plan depends on the exam date

Expect the spaced plan to feel slower. In the opening study, most participants judged massing the better method after their first session, even though spacing worked better for most of them.¹ When a returning topic feels rusty, that effort is the point of the gap.

3. Test yourself with practice questions from the start

Testing yourself should begin in the first week, before you have finished reading. In Roediger and Karpicke’s 2006 experiments with US undergraduates, rereading a passage won on a test given minutes afterward, but trying to recall it won two days and a week later, even though the recallers were never shown the right answers.⁷ What helps on the day you study is not what helps on the day of the exam.

Myth

Rereading your notes until they feel familiar is the safest way to prepare.

Fact

In 2006 experiments, rereading did better on a test minutes later, but recalling from memory did better days later.

Dunlosky’s review rated this retrieval practiceretrieval practice: Learning by recalling material from memory, for example by quizzing yourself, instead of rereading it. Pulling information out makes it easier to recall later, a result also known as the testing effect.Full entry in the glossary highly because its benefit showed up with many kinds of questions and material, in learners of different ages and over short and long delays, and because it is not especially time-consuming and needs little training.³ The details of formats and checking answers are covered in how to test yourself with active recall.

In the plan, that means opening each session with questions on the topics due that day, then using the misses to decide what to reread. A law student coming back to contract formation, for instance, might write short answers to a few problem questions from memory, then open the notes only for the rules she got wrong. As the exam gets closer, switch from short sets to past papers under a time limit, so you also practice the pace of the real thing; both the Auckland and Bath guides recommend timed practice with past questions.⁴⁵

4. Mix topic types once you can do each one

Mixing different types of problem in one session, known as interleaving, can help once each type is familiar, though its evidence is thinner than for spacing or self-testing. Dunlosky’s review rated it moderately useful: it has had large effects on learning math skills, but the body of research was small and included enough null results to raise concern, perhaps because some students had not practiced each task enough on its own first.³

Exam questions rarely tell you which chapter they come from. A statistics paper does not say “use a chi-square test here”, and a law paper does not label which area of law a problem scenario tests. Practice that mixes topics makes you do that choosing, which blocked practice does for you.

The practical rule is to keep a new topic in its own block until you can answer its questions without help, then shuffle. In the middle weeks, build practice sets from several topics at once, and let past papers do the mixing in the final weeks.

5. Use the last days for review, not new material

The last days before an exam are for strengthening what you already know. The University of Bath advises against trying to learn new things at that point, suggests using your notes and flashcards to trigger your memory, and warns against staying up all night revising.⁵ Singapore Management University suggests keeping a list of frequently tested or weak points to go through one more time right before the exam.⁶ For a medical or engineering candidate, that list might be a single page of the formulas, reference values or thresholds they most often mix up.

This step leans on university guidance rather than trials, but it fits the rest of the plan: if the earlier weeks worked, the final days are one more spaced return. Whether a last-minute push is ever worth it, and how to spend the night before, are questions of their own. If your worry is going blank on the day, the research on why minds go blank under exam pressure covers what helps.

Those topics go into the next session.

Your exam plan, working back from the date

• Exam date, format and question weightings written down
• Every syllabus topic listed and rated by a short quiz, not by feel
• A first pass for each topic scheduled early
• Each topic booked for at least two more visits on different days
• Every session opens with questions on the topics due that day
• Mixed practice sets once each topic can be answered alone
• Timed past papers in the final weeks
• Last days kept for review, with no new topics

Download PDF ↓

Get this checklist as a one-page PDF. We'll email you the download link.

Your emailAlso send me the weekly newsletter (launching soon)Email me the PDF

We use your address to send this link and, only if you tick the box, the newsletter. Privacy policy.

Trial-backed steps and adviser-backed steps

Spacing sessions and testing yourself have the firmest support, from large reviews and many experiments; mixing topics has fewer studies behind it, and mapping the exam and the final review rest on university guidance. No source here tested a complete exam plan against another plan, so the order of steps is our reading of the evidence, not a tested result.

Most of the experiments used students learning word pairs, facts or short passages, and the effects on a whole course exam are less well measured. Treat the steps as well-supported defaults and adjust the details to your exam.

The bottom line

Put the exam date on the calendar first and plan backward from it, early enough that every topic gets a first pass and at least two returns on different days. Open each session with questions rather than your notes, mix topics once each one is familiar, and save the final days for review. When you are unsure how long to wait before returning to a topic, wait longer.