Spaced Repetition Schedule: How Often to Review

A spaced repetition schedule spreads your reviews across growing gaps instead of one long session: review the material on day 1, day 3, day 7, day 14, then day 30. Each successful recall pushes the next review further out, so the same fact takes longer to fade.

What is spaced repetition?

Spaced repetition is reviewing material right before you are likely to forget it, then waiting longer before the next review, and longer again after that. The gaps grow because your memory for that item gets stronger every time you retrieve it successfully.

It runs on one mechanism. Retrieving a memory takes effort, and that effort is what strengthens the trace. Rereading skips the effort, which is why it feels easier and does less.

Timing is what turns the mechanism into a method. Retrieval is only hard, and therefore only useful, once the memory has started to fade. Everything below is about that timing: how long to wait, what to do when the plan slips, and what deserves a place in the cycle at all.

Why do you forget so fast?

Because forgetting is the default. Your brain deprioritizes information it does not use, and the drop is steepest right after you learn something, then slows down.

That shape has a name. In the 1880s, Hermann Ebbinghaus memorized lists of nonsense syllables and relearned them after gaps running from twenty minutes to a month, measuring how much of the original work he was spared the second time. In his own account of the results, an hour on, relearning a list still took about half the original effort. A day on, roughly a third of that effort was spared. After six days, about a quarter. After a month, a fifth. Plot those points against time and they fall steeply, then flatten into a long shallow tail. That is the forgetting curve.

Three things it does not say, and that most summaries get wrong:

• It measures effort saved, not answers recalled. Ebbinghaus tracked how much relearning he was spared, which is not the same quantity as how many syllables he could produce. When Murre and Dros replicated the experiment in 2015, their recall curves came out far flatter than their savings curves.
• It is one man, on meaningless material. The syllables were chosen precisely because they carried no meaning. His numbers describe that task, not a chapter you understood.
• It describes material that is never reviewed. That is the whole point of it. The curve is the baseline you are trying to beat, not a forecast of your grade.

The replication turned up one detail worth keeping: the curve is not smooth. It steps back up around the twenty four hour mark, plausibly because of sleep, though the authors are careful not to claim that outright.

What you should take from all of it is the timing, not the numbers. The first day after you learn something is the highest risk window, so the first review belongs there, not a week later when you are relearning from scratch.

What spaced repetition schedule should you follow?

Start with day 1, 3, 7, 14, 30, counted from the day you first covered the material. Nothing is magic about those numbers. They are a set of gaps that roughly doubles, which keeps each review landing near the edge of forgetting.

• Day 1, the same evening. A quick pass a few hours after class, while the material is still half there. This is the highest value review of the cycle, and the one most people skip.
• Day 3. The first real test. Whatever you cannot produce here is what you did not learn, as opposed to what you recognized in your notes.
• Day 7. The easy facts should feel automatic. Spend the session on the two or three that do not.
• Day 14. A shorter pass, mostly confirmation. If it takes long, day 7 went too easy on you.
• Day 30. A final check before an exam, or before moving on to material that builds on this one.

The gaps grow for a reason. A fixed interval, say every three days, spends effort on what you already know and still arrives too late for what you do not. Growing intervals track the memory instead of the calendar.

What if you miss a review?

Drop back one step rather than skipping ahead. Miss day 7 and reach it on day 12, and it counts as the day 7 review: put the next one at day 19, not day 21.

Missing several in a row is not a reason to start over either. Test yourself once. What you can still produce stays on schedule, what you cannot goes back to day 1. The schedule is only a default, and what you can actually recall overrides it.

How do you know an interval was too short or too long?

By how the recall felt, not by the date. Two signals, both reliable:

• Too short: the answer arrives instantly, with no effort at all. You spent a review for nothing. Push the next gap further out.
• Too long: you cannot produce the answer and have to read it again. You are relearning, which costs nearly as much as learning it the first time. Pull the next gap in.

What you want is the uncomfortable middle, where the answer arrives but takes a few seconds and some work. That effort is the mechanism, not a sign that something has gone wrong.

How long does it take each day?

Less than you expect, and less as the weeks pass. A day of reviews for one course usually runs ten to twenty minutes, because most of what is due is material you have already recalled correctly two or three times.

The load is heaviest in the first fortnight of a new topic, when everything sits at day 1 or day 3. It thins as items move to the 14 and 30 day gaps. That is the exact opposite of cramming, where the work piles up as the exam gets closer.

One practical consequence: start the cycle the week you cover the material, not the week before the exam. Starting late does not give you a lighter version of the schedule. It gives you cramming with extra steps.

What should go into the cycle, and what should not?

Discrete facts with one correct answer: definitions, vocabulary, formulas, dates, mechanisms, classifications. Anything you either know or do not know is a good candidate, and flashcards are the natural container for it.

What does not belong: whole chapters, essay arguments, and anything whose honest answer is "it depends". You cannot grade yourself fairly on a card that asks you to explain the causes of a war. Split it into the pieces that do have answers, the dates, the actors, the treaties, and practice the rest by writing or by solving problems.

Understanding is not exempt from review, it just needs a different exercise. In a subject built on reasoning, the spaced item is a problem to solve rather than a fact to retrieve. Same schedule, different task. And if a topic resists being turned into questions at all, that usually means you have not understood it yet, which active recall will expose faster than another read of the notes.

Does it still work if your exam is in five days?

Partly. You cannot run a 30 day cycle in five days, but the two things that make it work still apply: spread the passes, and retrieve instead of rereading.

On five days, a workable version is a pass today, one in two days, one the day before, and a short check on the morning itself. The gaps are compressed, so more of the material will still feel fresh at each pass. Compensate by being stricter: if you can only produce it with the notes open, it does not count.

What you should not do is turn the panic into one ten hour session. Two spread sessions of three hours beat a single six hour block, even on a five day horizon. If that is where you are, revising without cramming goes into the short horizon case in more detail.

Do you need an app, or is a calendar enough?

A calendar is enough up to a point, and that point is volume. One course and forty facts, writing 1, 3, 7, 14, 30 next to each topic works fine. Four courses and six hundred items, each on its own interval, and the bookkeeping becomes the reason you quit.

That is the honest argument for software: it tracks per item intervals so you do not have to. Tools built on spaced repetition, Orbily included, schedule each card separately, so the ones you keep missing come back sooner while the ones you know drift toward longer gaps. The flashcards feature page shows what that looks like in practice.

The method does not depend on the tool. Paper cards in five labeled boxes, one per interval, implement the same schedule, and did so long before any of it was software.

Spaced repetition vs cramming

Cramming, or massed practice, packs the review into one session close to the deadline. It genuinely works for the deadline: short term recall responds well to repetition packed tightly together, which is why the habit survives.

The cost lands afterwards. Material learned that way decays quickly, often within days, because nothing ever reset the forgetting curve. You relearn it from near zero the next time you need it, and that is the hidden bill: not the night you lost, but every hour spent re-covering the same ground next term.

What is solid, on students and real course material: at the same total study time, spacing wins on long term retention. Kapler, Weston and Wiseheart (2015) had 169 undergraduates review a simulated lecture either one day or eight days later, then tested them five weeks on. The spaced group came out ahead, including on application questions rather than only on factual recall.

You will also see the stronger claim that spacing takes less total time for the same result. That one is not established for coursework. The nearest evidence sits in other domains: Bahrick and colleagues (1993) found thirteen relearning sessions spaced 56 days apart gave retention comparable to twenty six sessions spaced 14 days apart, half the sessions for the same outcome, but that was four people learning foreign vocabulary across nine years, and the wider gaps slowed the initial learning. Suggestive, not a promise about your semester.

For anything you will need again, a final, a licensing exam, a skill for work, the trade is worth making regardless.

Where to start tonight

Take whatever you covered today, close the notes, and write down what you remember. That is your day 1 review and it costs ten minutes. Then put day 3, 7, 14 and 30 in your calendar for the same material. The schedule only has to survive its first week. After that, the growing gaps do most of the work for you.

Spacing is one piece of a working method rather than the whole of it. If you are rebuilding how you study from the ground up, a simple method that works puts this schedule back in context alongside the rest.