Spaced Repetition

Here is the experience every student knows: you study a chapter on Tuesday, feel solid on Wednesday, and by the following Monday it has mostly evaporated. The notes are still in your binder. The knowledge is not in your head. Most people respond by studying harder, which usually means re-reading the same chapter in one long sitting and forgetting it again on the same schedule.

Spaced repetition is the fix, and it is one of the most consistently replicated findings in the psychology of learning. The idea is almost insultingly simple: instead of reviewing material many times in one night, you review it a few times across many days, with growing gaps between each review. Same total effort, dramatically different retention. Cepeda et al. (2006) pooled hundreds of comparisons of spaced versus massed practice and found the spaced version winning across ages, materials, and retention intervals.

This article covers what spaced repetition actually is, why it works, and the part most guides skip: how to run it when you have a fixed exam date staring at you from the calendar.

The problem: memory decays on a curve

In 1885, Hermann Ebbinghaus memorized lists of nonsense syllables and then tested himself at intervals from twenty minutes to a month. What he found became the forgetting curve: retention drops steeply in the first hours and days after learning, then the decline levels off. The shape has held up remarkably well in modern replications.

Two things about that curve matter for studying.

First, the steep early drop means that a single exposure to material, however attentive, is mostly a sunk cost. A lecture you understood perfectly on Tuesday is not “learned” in any durable sense. It is learned the way a sandcastle is built.

Second, and this is the part that makes spaced repetition possible: every time you successfully recall something, the curve resets and flattens. After the first review you might hold the material for days instead of hours. After the second, weeks. The decay never stops entirely, but it slows enough that a handful of well-timed reviews can carry a fact to exam day and well beyond.

What spaced repetition actually is

Spaced repetition is a review schedule built on that resetting behavior. You review each item at expanding intervals, with each review placed roughly where the forgetting curve predicts you are about to lose it.

A typical sequence for one fact looks like this:

Review	When	What it feels like
Learn	Day 0	Effortful, new
1st review	Day 1	Shaky but recoverable
2nd review	Day 3	Some hesitation
3rd review	Day 7	Mostly smooth
4th review	Day 14–21	Quick confirmation

The exact numbers are not sacred. What matters is the shape: short gap first, then progressively longer ones, with the next interval depending on how the last review went. Answer easily and the gap stretches. Fumble and the item drops back to a short interval until it stabilizes.

Note what this is not. It is not reviewing your whole deck every night, which wastes time on things you already know. And it is not a flashcard-only technique. You can space anything you can test yourself on: flashcards, practice problems, a blurting session on an essay topic, even re-deriving a proof.

You also do not need software to run it, just honesty and a bit of bookkeeping. The classic paper version is the Leitner box: index cards sorted into a few compartments, where correctly answered cards advance to a compartment reviewed less often and missed cards drop back to the daily one. It works. The reason most students eventually automate is volume: with three courses and a few hundred items, each on its own clock, the sorting becomes a part-time job, and a skipped sorting day quietly corrupts the whole schedule.

Why it works: two effects stacked

Spaced repetition gets its power from two well-documented effects working together.

The spacing effect

Information studied across separated sessions is retained far better than the same information studied for the same total time in one session. This was already visible in Ebbinghaus’s data, and the Cepeda et al. (2006) meta-analysis confirmed it at scale. One plausible explanation: when you return to material after partially forgetting it, your brain has to do real reconstructive work, and that work strengthens the memory in a way that effortless same-day repetition cannot.

This is Robert Bjork’s idea of “desirable difficulties” in action. The mild struggle of recalling something half-forgotten is not a sign the method is failing. The struggle is the method.

The testing effect

The second ingredient is retrieval. Roediger and Karpicke (2006) had students either repeatedly re-read a passage or read it once and then repeatedly recall it from memory. On a test a week later, the recall group decisively outperformed the re-readers, even though the re-readers had spent more time with the text in front of them. Pulling information out of memory strengthens it more than putting it in again.

So a spaced repetition session should always be a test, never a re-read. The full case for this is in our piece on active recall, but the short version: if your “review” consists of looking at the answer, you are running the schedule without the engine.

Why a little forgetting is part of the design

Bjork’s framework distinguishes two properties of a memory: how well it is stored, and how easily it can be retrieved right now. Cramming inflates the second while barely touching the first, which is why crammed material feels solid at midnight and is gone by the weekend. Spaced retrieval does the opposite. By waiting until an item has become slightly hard to reach, each review forces a deeper reconstruction, and that effort is what builds lasting storage. Counterintuitive consequence: the schedule is supposed to let you start forgetting. If every review feels easy, your intervals are too short and you are paying for retention you would have had for free.

Running it toward a fixed exam date

Most spaced repetition advice assumes you are learning a language over five years. Students do not have five years. You have a pathophysiology exam on the 14th, and the schedule has to guarantee every item gets enough reviews before that date.

That constraint changes the math in three ways.

Front-load new material. Every new item needs three to five reviews at expanding intervals to stick. Working backwards, an item introduced four days before the exam can only ever get cramped, short-gap reviews. So the rule is: finish converting your notes into questions early, ideally in the first third of your prep window, and spend the rest of the time reviewing. A study schedule that says “learn chapters 1–8 in week one, review weeks two and three” will beat “one chapter per day until the exam” every time.

Compress the intervals to fit. With a three-week runway, the classic 1–3–7–21 spacing does not fit for material learned in week two. Intervals have to shrink proportionally: maybe 1–2–4–7. You lose a little long-term durability, but everything still gets multiple spaced retrievals before test day. This is also where doing it by hand gets genuinely annoying, because every item is on its own clock. StudyDone handles this automatically: it turns your notes or slides into questions, then paces each item’s reviews to your exam date, so the things you keep missing show up more often and the things you know stay out of your way.

Plan a final consolidation pass. In the last two or three days, the queue should naturally surface mostly your weak items. Let it. Resist the urge to comfort-review the easy ones. The night before, do one fast pass through recent misses and stop; a well-run queue leaves very little for that night to do.

A concrete example

Say you have a 60-slide pharmacology deck and 18 days. A workable plan:

1. Days 1–5: convert the slides into question-and-answer items, about 12 slides per day, and do the day-1 review of each batch. A flashcard generator that works from your slides collapses this step from hours to minutes.
2. Days 6–15: no new material. Just clear the daily review queue, which will run 15–30 minutes most days.
3. Days 16–18: the queue thins out on its own. Add one or two timed practice tests for exam-format realism.

Total daily commitment after the setup week: well under an hour. That is the trade spaced repetition offers, and it is a good one: small, boring, daily payments instead of one catastrophic balloon payment the night before.

Common mistakes

The method fails for predictable reasons, and almost never because “spaced repetition doesn’t work for me.”

• Recognition cards. A card you answer with “yeah, I’d recognize that” is not testing recall. Every item should force you to produce something from memory before you flip it. Our guide to making good flashcards covers the card-writing side in detail.
• Skipping days. The schedule’s whole value is timing. A skipped day means reviews land after the forgetting has already happened, which turns cheap maintenance into expensive relearning. Ten minutes on a bad day preserves the system.
• Reviewing what feels good. Left to choose, everyone re-reviews the cards they already know, because fluency feels like progress. Trust the queue. The uncomfortable cards are where the marks are.
• Marking “got it” too generously. If you hesitated, needed the first letter, or got it half right, that is a miss. Grading yourself softly inflates intervals and the item collapses right before the exam.
• Starting too late, then abandoning ship. Even five days out, spacing beats massing with the same hours. Compressed spaced repetition is still spaced repetition.
• Adding new cards forever. Every new item creates future review debt. If your daily queue has crept past what you can finish, the fix is to stop adding, not to stop reviewing. Near an exam, a smaller set of questions you actually see twice beats an enormous deck you abandon.

A note on the obvious objection: plenty of students pass exams by cramming, so why bother? Because the comparison isn’t pass versus fail, it’s cost versus durability. Cramming buys short-term retrieval at full price and forfeits the material within weeks, which matters a great deal in cumulative subjects, where this term’s pharmacology is next term’s prerequisite. Spacing costs the same hours, spread thinner, and leaves knowledge that survives into the next course and the licensing exam after that.

Putting it into practice

If you want to start tonight, here is the minimal version. Take whatever you studied today and write ten questions about it, answers hidden. Test yourself tomorrow, then three days later, then a week later, adding new questions each day you study. Grade yourself honestly and let the misses come back sooner.

That is the entire system. Everything else, from algorithms to apps, is just automation of those four sentences. For where spaced repetition sits among everything else that works, see our ranked review of the best study methods backed by science; the short answer is that it shares the top tier with retrieval practice, and the two are best used as one habit.

FAQ

How long before an exam should I start spaced repetition?

Two to four weeks works for most courses, because it leaves room for three to five reviews per item. With less time the schedule compresses: intervals shrink and daily sessions get longer, but the method still beats massed cramming with the same hours.

Does spaced repetition work for math and problem-solving subjects?

Yes, if you space whole problems rather than isolated facts. Re-solve a problem type from scratch a few days after you first learned it, then again the following week. The skill you are spacing is choosing and executing the method, not reciting a formula.

How many reviews does each fact need before an exam?

Most items are solid after three to five successful retrievals spread over expanding intervals. Hard items naturally earn more reviews because you keep missing them, which is exactly how the schedule should allocate your time.

Is spaced repetition the same thing as using flashcards?

No. Flashcards are a format; spaced repetition is a schedule. You can space anything you can test yourself on, including practice problems and blurting sessions, and you can use flashcards badly by reviewing the whole deck every night.

What happens if I miss a few days of reviews?

You forget some of what was due, and the backlog grows, but nothing is ruined. Clear the overdue queue over two or three sessions, expect to miss more cards than usual, and let the misses reset to shorter intervals.