Leitner System: How the Flashcard Box Method Works

The Leitner system sorts flashcards into boxes with different review intervals. Every card starts in the first box. A correct answer moves the card to a box reviewed less often, while a wrong answer returns it to the first box. Difficult cards therefore come back sooner without requiring a separate calendar calculation for every answer.

The method is simple enough to run with index cards and three envelopes. Most confusion comes from the calendar because guides use different box counts, intervals, and failure rules. Pick a clear version and keep its rules consistent.

How the Leitner system works

Each card has a question or prompt on the front and a short answer on the back. Put every new card in Box 1. When a box is due, attempt each answer before turning the card over.

• Correct in Box 1: move the card to Box 2.
• Correct in Box 2: move the card to Box 3.
• Correct in the final box: keep it there or retire it after a separate long review.
• Incorrect in any box: return the card to Box 1.

Each box has a longer interval than the one before it. A simple version reviews Box 1 after one day, Box 2 after three days, and Box 3 after seven days. Other versions use calendar days such as Monday, Wednesday, and Saturday. The University of York’s guide gives daily, every-other-day, and weekly reviews as one example.

The movement rule handles two jobs. A flashcard makes you retrieve one answer without looking. The box decides when that card returns. Correct answers earn a longer gap, while wrong answers shorten it. A weak card stays in frequent circulation and a familiar card stops taking up daily study time.

If you want more options for choosing intervals, the broader spaced repetition schedule guide covers fixed schedules and algorithmic approaches. A Leitner deck is the low-tech version with rules you can see on your desk.

What the evidence supports

The research basis is stronger for the ingredients than for one exact box arrangement. A major 2013 review of learning techniques rated practice testing and distributed practice as high-utility methods. Practice testing means attempting an answer from memory. Distributed practice means spreading learning across time instead of packing it into one session.

The review summarized 254 spacing studies with more than 14,000 participants. Average recall was 47% after spaced study and 37% after massed study. That is useful evidence for spacing. It does not prove that three boxes beat five, or that a 1-day, 3-day, 7-day calendar is ideal for every subject.

The same review notes that the evidence is thinner for complex, structured knowledge and tasks requiring higher-level thinking. A card can test a definition cleanly. It is less suited to judging whether you can critique an argument, design an experiment, or debug a program. The method helps most when the material can be written as a clear prompt with a checkable answer.

A complete three-box example

Here is a six-card deck for basic statistics. The schedule uses Box 1 after one day, Box 2 after three days, and Box 3 after seven days. A correct answer moves forward one box. A wrong answer returns to Box 1. The next date is counted from the review that just happened.

The results are scripted to show the mechanics. They are not evidence that somebody learned statistics in two weeks.

Card	Prompt
A	What is the mean?
B	What is the median?
C	What is the mode?
D	What is the range?
E	What does standard deviation describe?
F	What does correlation measure?

All six cards start in Box 1 and are due on day 1.

Day	Cards reviewed	What happens
1	A, B, C, D, E, F	A, C, D, and F move to Box 2, due day 4. B and E fail and stay in Box 1, due day 2.
2	B, E	B moves to Box 2, due day 5. E fails again and is due day 3.
3	E	E moves to Box 2, due day 6.
4	A, C, D, F	A, D, and F move to Box 3, due day 11. C fails and returns to Box 1, due day 5.
5	B, C	B moves to Box 3, due day 12. C moves to Box 2, due day 8.
6	E	E moves to Box 3, due day 13.
8	C	C moves to Box 3, due day 15.
11	A, D, F	A and F stay in Box 3, due day 18. D fails and returns to Box 1, due day 12.
12	B, D	B stays in Box 3, due day 19. D moves to Box 2, due day 15.
13	E	E fails and returns to Box 1, due day 14.
14	E	E moves to Box 2, due day 17.

Days 7, 9, and 10 have nothing due. By the end of day 14, Box 1 is empty, Box 2 contains D and E, and Box 3 contains A, B, C, and F. The full trace contains 24 reviews. Every card appears on its assigned day, and the deck still contains the same six cards.

Notice what the boxes reveal. The scripted log marks E incorrect three times, so it receives more attempts and two fresh runs through Box 1. A and F follow the slow path with no resets. The review load changes in response to answers rather than the order in which the cards were written.

How to set up your own Leitner box

Three envelopes are enough for a first deck. Label them with the interval rather than vague confidence terms. 1 day, 3 days, and 7 days are harder to misread than hard, medium, and easy.

Write one answerable prompt per card. A useful card asks for a definition, distinction, step, label, formula, or short explanation. If the back needs half a page, split the card or choose a different review format. The active recall guide has several ways to test broader material without forcing it into tiny cards.

Keep a due date on each card or divide each box by date. Reviewing an entire higher box every Tuesday can cause a card promoted on Monday to appear again one day later. A visible due date preserves the interval you chose.

Start with a small deck. Ten good cards will teach you more about the workflow than 200 cards copied from a chapter and quietly avoided.

Common failure modes

The card is too broad

“Explain statistics” cannot produce a clean right or wrong decision. “What is the median?” can. Rewrite cards that fail because the prompt is muddy rather than because the answer was forgotten.

You change the rule during review

Moving a nearly correct card forward feels merciful, but it makes the boxes meaningless. Decide whether failure means returning to Box 1 or moving back one box, then use that rule for the whole deck.

Box 1 becomes a paper avalanche

Every new and failed card returns to the fastest interval. Adding a large chapter in one afternoon creates a large daily review tomorrow. Pause new cards when Box 1 becomes too slow to finish.

You treat the final box as permanent mastery

The final box means the card survived the current schedule. It does not mean the answer is permanent or useful in every context. Keep a long review interval, run an occasional mixed check, or retire the card only when forgetting it would no longer matter.

When boxes are the wrong review unit

The Leitner system is good at discrete material: vocabulary, anatomy labels, dates, symbols, definitions, and short procedures. It becomes awkward when the useful unit is a whole book chapter, lecture, project, or technical concept.

Suppose you want to retain a chapter about statistical inference. Turning every paragraph into cards may create more maintenance than value. A whole-topic review could instead ask you to explain the argument, reproduce the main diagram, solve one representative problem, and check the gaps against your notes.

This keeps retrieval on a schedule but uses a larger unit. The learner judges the quality of the answer.

Where Spacey fits

Spacey is a topic-level review scheduler for iPhone and iPad. You could add Basic statistics as one study and use the Short plan, which schedules reviews on days 1, 3, and 7. At each review, you choose the useful action, such as explaining the six terms from memory and checking your notes.

Spacey is not a digital Leitner box. It does not hold the six cards, notice that standard deviation failed, or send that one answer back to tomorrow. It remembers when the topic returns. You decide what happens inside the review.

Use a Leitner deck when individual answers deserve individual schedules. Use topic-level scheduling when the material makes more sense as a connected whole. A shoebox is surprisingly competent, but it should not get to choose the shape of everything you learn.