Paced.›Guides

Spaced repetition schedules compared: which intervals, from where, at what cost

Published 23 October 2026 · Updated 23 October 2026

Sources checked on 23 October 2026 by Alexandre Doumergue, founder of Paced.

The schedules at a glance

Disclosure: Paced is our product. Its paces are listed for completeness, in the last row, not ranked. Where the source column says “no primary study found”, we searched PubMed titles and abstracts for the name of the rule on 23 October 2026 and found no record.

Fixed schedules

Pimsleur's graduated intervals (1967)

Paul Pimsleur published “A memory schedule” in The Modern Language Journal in 1967 (Pimsleur, 1967), an early proposal of recall gaps that grow, made for language learning. We checked its bibliographic record but could not read the article, so this page quotes none of its figures; the sequence you may see attributed to it online is not repeated here. What it shares with every later schedule is the growing gap.

1-3-7-14-30, 2-7-30 and 2-3-5-7

These three sequences circulate on study-skills pages, apps and social media. We found no study that tested any of them as such. They are reasonable because the gaps grow; they are limited because they stop — after 30 days, 30 days and 7 days respectively. For material you need beyond that, add longer gaps. The 2-3-5-7 method has its own page, with the three readings of its numbers and a calculator.

The French “méthode des J”

French medical students plan their courses with “J” days: J0, J1, J3, J7, J15, J30, J60, J120, with J0 the day of the first study and a review on each later J. Paced's Standard pace uses the same days and goes on to day 240. In French, the method has its own page on pacedstudy.fr.

Adaptive schedules

The Leitner system

Sebastian Leitner described a box of flashcards divided into decks in his book So lernt man lernen. In the version analysed by Reddy et al. (2016), a card enters the first deck; each deck is a queue; when you review a card you got right it moves up one deck, and a card you got wrong goes down one. The aim is that lower decks are reviewed more often than higher ones. The gaps are therefore not fixed in days: they come from how often you open each deck, and Reddy et al. note that existing schemes for setting those frequencies rest on heuristics.

SM-2 (SuperMemo)

SM-2 is the algorithm Piotr Wozniak described for SuperMemo: the first gap is 1 day, the second 6 days, and each later gap is the previous one multiplied by an E-Factor that starts at 2.5, moves with the grade you give (0 to 5), and never goes below 1.3; a fractional gap is rounded up (Wozniak, 1990). If every answer is graded 4 (“correct response after a hesitation”), the E-Factor stays at 2.5 and the reviews fall on D+1, D+7, D+22, D+60, D+155, D+393. That counts each rounded gap as the base of the next; the text does not say whether to round before or after, and with the unrounded chain the last two reviews fall on D+154, D+389. Anki's older scheduler is based on SM-2.

FSRS

FSRS (Free Spaced Repetition Scheduler) gives each card a memory state of three values — retrievability, stability and difficulty — estimated from your review history, and schedules the next review from it. The Anki manual describes it as an alternative to the older SM-2-based algorithm, available since Anki 23.10 (Anki manual). It needs a history of reviews per card, which is why it suits large card collections and not a list of ten topics.

What each schedule costs

Fixed schedules can be costed in advance. For a topic first studied for 2 h 30, with Paced's default review lengths (30% of the first study for the first review, four points less for each next one, never under 10%), our calculation:

The time per week that follows from these numbers is in the spaced repetition calculator, which draws it for a pace and an intake of topics.

What is the best interval?

Searches such as “spaced repetition intervals”, “spaced repetition best intervals” and “spaced repetition how many days” ask the same thing: how long to wait before the next review.

The best-supported answer is a relation, not a number. In a study of more than 1,350 learners, the best gap grew with the time to the test (about 20–40% of it for a test a week away, 5–10% for a test a year away) (Cepeda et al., 2008). A meta-analysis of 839 assessments from 317 experiments found that the gap that maximises retention also grows with how long you need to remember (Cepeda et al., 2006). Our reading of that relation, not a rule from the studies:

• exam in one to two weeks: short gaps (2-3-5-7, or the Intensive pace);
• exam in one to three months: growing gaps into weeks (1-3-7-14-30, or the Standard pace);
• knowledge you need long after the exam: gaps into months (the Light pace, SM-2, FSRS).

Anki is a trademark of Ankitects Pty Ltd. Paced is not affiliated with, or endorsed by, Anki or Ankitects.

SuperMemo is a trademark of SuperMemo World. Paced is not affiliated with SuperMemo World.

Questions

Sources

1. Cepeda, N. J., Vul, E., Rohrer, D., Wixted, J. T., & Pashler, H. (2008). Spacing effects in learning: A temporal ridgeline of optimal retention. Psychological Science, 19(11), 1095–1102. More than 1,350 learners, review gaps up to 3.5 months, final tests up to a year later.
2. Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380. Meta-analysis of 839 assessments in 317 experiments: the gap that maximises retention grows with how long you need to remember.
3. Wozniak, P. A. (1990). Application of a computer to improve the results obtained in working with the SuperMemo method. Optimization of learning, Master's thesis, University of Technology in Poznan; web version on super-memory.com. The SM-2 algorithm: a first gap of 1 day, a second of 6, then each gap is the previous one times an E-Factor that starts at 2.5 and never falls under 1.3; a fractional gap is rounded up.
4. Reddy, S., Labutov, I., Banerjee, S., & Joachims, T. (2016). Unbounded human learning: Optimal scheduling for spaced repetition. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 1815–1824. Describes the Leitner system (decks of cards; a recalled card moves up one deck, a missed one down one) and cites Leitner's book So lernt man lernen (Herder, 1974).
5. Ankitects (Anki manual) (2026). What spaced repetition algorithm does Anki use?. docs.ankiweb.net. Since Anki 23.10, two algorithms: one based on SM-2, and FSRS, which gives each card a memory state (retrievability, stability, difficulty) estimated from the review history.
6. Pimsleur, P. (1967). A memory schedule. The Modern Language Journal, 51(2), 73–75. An early proposal of recall gaps that grow, made for language learning. Only its bibliographic record was checked; the page gives none of its figures.

Read next

• The 2-3-5-7 revision method: what it is, what it misses, and a calculator

  2-3-5-7 method

• Spaced repetition schedule calculator: the date and length of every review

  Schedule calculator

• The forgetting curve: what it actually shows, and how to schedule reviews around it

  The forgetting curve

• Spaced repetition: which intervals, and how to plan them

  Spaced repetition