How to Study for an Exam: A Spaced Repetition Schedule Based on Research
Published: September 28, 2026 | By: Roman Vasylyshyn, founder of Lambdio | Reading time: 9 min
Most students study for an exam the same way: a few light sessions early on, a growing sense of dread, and then an intense burst in the last 24 to 48 hours. It feels like the natural rhythm of exam prep. It is also close to the least efficient way to turn study hours into exam points.
The research on this is unusually clear. The same number of study hours, spread out in the right pattern, produces better recall on exam day than the same hours packed into a cram session. This post covers how far ahead to start, how to space your sessions, what to do in each session, and what to do if you've already left it too late.
Table of Contents
- The Short Answer
- Why the Timeline Matters More Than the Hours
- How Far Apart Should Sessions Be?
- A Sample Schedule for a 4-Week Exam Timeline
- What to Do Inside Each Session
- If You're Starting Late
- Common Mistakes
- How Lambdio Handles the Scheduling For You
The Short Answer
- Start 3-4 weeks before the exam if you can, for material you need to know well.
- Break the content into chapters or topics and revisit each one several times, rather than finishing one topic and never returning to it.
- Space your reviews out progressively: short gaps at first, longer gaps as the material sticks.
- Use active recall in every session (quizzing yourself, explaining out loud) rather than rereading.
- Save the last 2-3 days for retrieval practice on your weakest material, not for first-time learning.
The rest of this post explains why each of those points holds up.
Why the Timeline Matters More Than the Hours
Two mechanisms we've covered in earlier posts drive everything below.
The first is the forgetting curve: newly learned material decays quickly, but each well-timed review flattens the curve and makes the next interval longer. A schedule built around that pattern needs fewer total hours to reach the same retention.
The second is the testing effect: retrieving material from memory strengthens it far more than rereading it does. So the best schedule isn't just about when you study, it's about making every session a retrieval event.
Together, these mean two students can put in the same 20 hours and finish with very different exam outcomes, depending on whether those hours were massed into a few days or distributed across weeks, and whether they were spent rereading or retrieving.
How Far Apart Should Sessions Be?
This is the question that turns "spaced repetition works" into a usable plan, and it has a research-backed answer.
In 2008, Nicholas Cepeda and colleagues ran a large study (over 1,300 participants) that varied both the gap between study sessions and how long people had to retain the material afterward. Their finding: the optimal gap between study sessions depends on how long you need to remember the material. Gaps that are too short waste the benefit of spacing; gaps that are too long let the memory fade too far before the review.
The rough rule that emerges from the data is that the ideal gap is about 10-20% of the time until you need the information. Need to remember something for a week? Review it after roughly a day. Need it for a month? Review it after roughly 3-6 days. Need it for six months? The first gap is closer to 3-4 weeks.
For exam prep, this means your first review of a chapter should be a few days after you first learn it (for a 4-week timeline), not the same evening and not two weeks later. Subsequent reviews can stretch out further.
The pattern of retention this produces looks like this, compared with cramming the same number of sessions into the final days:
This chart is an illustration of the qualitative pattern the spacing literature describes (crammed material feels strong right after studying but drops off quickly, spaced material shows a slower decay from a higher base), not a measurement from one specific study. The specific numbers will vary by person, material, and how difficult the content is.
A Sample Schedule for a 4-Week Exam Timeline
Here's a concrete schedule for a course with roughly 8 chapters or topics, starting 28 days before the exam. Adjust the numbers to your own timeline, keeping the ratio of gap to time-remaining roughly consistent.
| When | What to do |
|---|---|
| Days 28-21 | Learn each chapter for the first time, one every day or two. End each session with a short self-quiz on what you just covered. |
| Days 21-14 | First review of each chapter, roughly 4-6 days after you learned it. Retrieval only: quiz yourself before opening your notes. |
| Days 14-7 | Second review of each chapter, with more attention on the ones you struggled to recall. |
| Days 7-3 | Third review, focused on your weakest chapters. Introduce practice exams or past papers under timed conditions. |
| Days 3-1 | Light retrieval practice on everything, with extra time on the weakest material. No new content. |
| Day before | A short, final self-quiz. Then sleep. Sleep is part of consolidation, not a cost to it. |
Notice that the last week contains no first-time learning. The timeline is deliberately front-loaded: everything you need to know should already have been seen at least once by the time you enter the final stretch, so that the last week is about strengthening memories rather than building them from scratch.
What to Do Inside Each Session
The schedule tells you when to study. What you do inside each session matters just as much, and the guidance from earlier posts translates into a simple structure:
- Start with retrieval, not review. Before opening your notes or the chapter, try to recall what you can. Write down, say out loud, or answer questions about the material from memory. Then check what you got wrong or missed.
- Explain the material in your own words. For anything you're unsure about, try explaining it as though to someone new to the subject. Where your explanation stalls is where your understanding has a gap. (More on why this works in our post on Socratic questioning and the Feynman Technique.)
- Prioritize by difficulty, not by comfort. It's natural to spend review time on material you already know well because it feels productive. The higher-value use of time is the material you keep getting wrong.
- Keep sessions focused. Several shorter sessions on a chapter usually beat one long one. There's no magic number, but 25-45 minutes of focused retrieval is a reasonable unit for most people.
- Finish by noting what to revisit. A quick list of "I still struggle with X and Y" turns into the priority list for your next session.
If You're Starting Late
Not everyone has four weeks. If you're starting with a week or less, the same principles apply in compressed form:
- Prioritize ruthlessly. With limited time, identify the material most likely to be tested and most likely to be forgotten, and start there.
- Compress the gaps, keep the structure. With a week to go, a rough pattern is: learn on day 1, review on day 2-3, review again on day 5, final pass on day 6-7. It's not as effective as a longer timeline, but it's still meaningfully better than massed studying.
- Do retrieval in every session, even a short one. Under time pressure, this is the highest-leverage habit to keep, because it makes each hour count for more.
- Accept the tradeoff. A compressed schedule produces good short-term recall, but it won't lay down the longer-lasting memory that a multi-week schedule does. That's a reasonable tradeoff if the exam is a few days out and you need to pass it, but it's worth knowing the difference if the material matters beyond the exam (for a cumulative course, for instance, or a professional certification).
Common Mistakes
- Waiting until the material "feels ready" to start reviewing. The point of spaced review is to catch a memory before it fades, which means reviewing while it still feels mostly fine.
- Treating rereading as review. It feels like review, and it isn't the same thing as retrieval.
- Reviewing only what you're already good at. Comfort-driven review inflates your confidence without improving your weakest areas.
- Skipping sleep for extra study. Sleep supports memory consolidation, and the cost of a lost night usually outweighs the benefit of the extra hours.
- Learning new material in the last 48 hours. It's better to spend that time strengthening what you already know than to add material you'll only half retain.
How Lambdio Handles the Scheduling For You
A schedule like the one above works, but it has one real weakness: it's static. It assumes every chapter is equally hard, that you'll remember each at the same rate, and that you'll follow the plan without slippage. In practice, some chapters stick easily and others don't, and life gets in the way.
This is what the FSRS scheduling engine in Lambdio is built to handle. Instead of a fixed calendar, each chapter has its own estimated difficulty and stability, updated after every session based on how you rated it. The system schedules the next review for the point at which your recall probability is predicted to fall to about 90%, so hard chapters come back sooner and easy chapters get pushed out further, automatically.
The learning modes map onto the in-session structure above: Standard and Socratic modes for first-time learning, Chat Review and Feynman mode for explaining and consolidating, and Quiz mode (which unlocks once the scheduler estimates a chapter is approaching the point of being forgotten) for retrieval practice at exactly the moment it does the most good.
The result is closer to a personal version of the schedule in this post, recalculated after every session rather than fixed on day one.
If you have an exam coming up and want to try it, courses are free to start at lambdio.com.
Sources: Cepeda, Vul, Rohrer, Wixted & Pashler (2008), "Spacing Effects in Learning: A Temporal Ridgeline of Optimal Retention," Psychological Science; Cepeda, Pashler, Vul, Wixted & Rohrer (2006), "Distributed Practice in Verbal Recall Tasks: A Review and Quantitative Synthesis," Psychological Bulletin; Roediger & Karpicke (2006), "Test-Enhanced Learning," Psychological Science.