Spaced repetitionSpaced repetition
for vocabulary

What it is, in plain words

You learn a new word today. Tomorrow you test yourself on it: cover the meaning and try to recall it. If you remember it, you wait a few days before the next test, then a week, then weeks. If you forget it, it comes back soon and climbs again. Over time the words you know take almost none of your time, and your minutes go to the words that are still slipping.

That is all spaced repetition is. The rest is bookkeeping: which words are due today, and how long each one should wait. Here is one schedule, the five boxes Lexpad uses by default:

Why it works

Spacing. In 1885 Hermann Ebbinghaus tested his own memory of lists of syllables and found that, in his words, “38 repetitions, distributed in a certain way over the three preceding days, had just as favorable an effect as 68 repetitions made on the day just previous” (Ebbinghaus, Memory, section 34). A review of 317 experiments later found the spacing effect again and again, and found that the right gap grows with how long you want to remember (Cepeda et al., 2006).

Testing. Recalling a word does more than looking at it again. Students who learned Swahili–English word pairs and kept testing themselves remembered far more a week later than students who kept studying the pairs (Karpicke and Roediger, 2008, Science). A review of ten study techniques rated these two, practice testing and spaced practice, the most useful (Dunlosky et al., 2013).

For vocabulary the two fit naturally: a flashcard is a test, and a schedule spaces the tests. How to memorize vocabulary covers the whole method, including keeping each word with its sentence.

Three kinds of schedule

Every spaced-repetition app has to decide when a word comes back. Most use one of three kinds of rule.

Leitner boxes are the oldest and simplest. Sebastian Leitner described them in his 1972 book So lernt man lernen: cards move up a box when you know them and back to the first box when you miss, and each box has its own wait. The Leitner system explained.

SM-2 is the rule Piotr Woźniak wrote for the first SuperMemo program in 1987: the first gap is 1 day, the second 6 days, and each later gap is the previous one times the card’s easiness factor, which starts at 2.5 and falls when a card is hard. A grade below 3 starts the card’s repetitions again (SuperMemo, algorithm SM-2). Anki’s older scheduler is SM-2.

FSRS (Free Spaced Repetition Scheduler) is open source. It keeps a small model of your memory of each card and schedules the review for the day your chance of recalling it falls to a target you choose. It grew out of research on memory data from the language app MaiMemo (Ye, Su and Cao, KDD 2022). In the open SRS benchmark, run on review histories from Anki users, it predicts recall better than many other models; that benchmark measures prediction and has no Leitner-box model.

Which apps use which schedule

“Which apps use FSRS?” is a common question, and the answer changes as apps add schedules. This table lists only what each app’s own documentation or listing said on the day we read it. Other apps may use these schedules too.

Making it work

• Answer honestly. Marking a word right when you were not sure only brings it back later, when it is harder.
• Test both ways. Word to meaning, then meaning to word. Recalling the word is harder, and it is what you need to use it.
• Keep sessions short and daily. A few minutes a day beats an hour once a week.
• Do not panic after a break. Nothing is lost; due words wait. In Lexpad, a daily limit on reviews (50, 100 or 200) spreads a backlog over several days.
• Use your own words. Words you met, with the sentence you met them in, are easier to keep than a list.

Sources

1. Ebbinghaus, H. (1885). Memory: A Contribution to Experimental Psychology, translated 1913, chapter 8, section 34. Classics in the History of Psychology.
2. Cepeda, N. J., et al. (2006). Distributed practice in verbal recall tasks. Psychological Bulletin 132(3). doi.
3. Karpicke, J. D., and Roediger, H. L. (2008). The critical importance of retrieval for learning. Science 319. doi.
4. Dunlosky, J., et al. (2013). Improving students’ learning with effective learning techniques. Psychological Science in the Public Interest 14(1). doi.
5. Leitner, S. (1972). So lernt man lernen. Freiburg: Herder.
6. Woźniak, P. (1990). Algorithm SM-2, from the first SuperMemo program (1987). SuperMemo archive.
7. Ye, J., Su, J., and Cao, Y. (2022). A stochastic shortest path algorithm for optimizing spaced repetition scheduling. KDD 2022. doi; open SRS benchmark.
8. App documentation, read 14 October 2026: Anki manual, Mochi docs, RemNote help, Quizlet on the US App Store, wokabulary.com.