Spaced repetition is a principle, not a fixed schedule. The intervals that work best depend on how long you want to retain the material, how difficult it is for you personally, and how much forgetting has occurred between reviews. But the research is precise enough to give us clear guidelines — and those guidelines are often counterintuitive.
What the research says: the 10–20% rule
Cepeda et al. (2006) conducted the largest meta-analysis of spaced practice research ever published: 254 studies, over 14,000 participants. Their key practical finding: the optimal gap between initial study and first review is approximately 10–20% of the target retention interval.
This means:
| Retention goal | Optimal first review |
|---|---|
| Remember for 1 week | 1–2 days after studying |
| Remember for 1 month | 3–6 days after studying |
| Remember for 3 months | 9–18 days after studying |
| Remember for 6 months | 18–36 days after studying |
| Remember for 1 year | 5–10 weeks after studying |
Review too soon (same day): the memory is still fresh, retrieval is easy, and the strengthening effect is minimal. Review too late: significant forgetting has occurred and retrieval often fails completely. The 10–20% window produces the desirable difficulty that maximises consolidation.
Building a practical review schedule
For most academic learners, a four-review schedule produces near-durable long-term retention for most material:
For a 6-month retention goal (typical exam or professional use):
| Review | Timing | Session format |
|---|---|---|
| Review 0 (initial study) | Day 1 | Read + take notes + brief free recall |
| Review 1 | Day 2–3 | Active recall only — retrieve without re-reading |
| Review 2 | Day 7–10 | Active recall — note items that fail |
| Review 3 | Day 21–28 | Active recall — focus on items from review 2 |
| Review 4 | Month 3 | Brief check — note what has faded |
Each review session should be retrieval practice — producing answers from memory — not passive re-reading. Re-reading is among the least effective study techniques (Dunlosky et al., 2013). The review is the practice of effortful retrieval.
The calendar approach: putting dates in your diary
The most reliable way to follow a review schedule is to schedule the review sessions in advance, the same day you study the material. This requires 30 seconds and prevents the most common failure mode: forgetting to review.
A simple system:
- Study new material
- Immediately open your calendar
- Add review events at: tomorrow, day 5, day 14, day 30
- At each review: retrieve from memory for 10–15 minutes, then check your notes
If a review reveals multiple failures, add an additional review event 3 days later.
How Anki schedules your reviews
Anki's SM-2 algorithm personalises intervals based on actual recall performance. The default schedule for a new card:
- Day 1: First review (after initial learning)
- Day 4: Second review (first good recall)
- Day 10: Third review
- Day 25: Fourth review
- Day 62: Fifth review
- Day 156: Sixth review
These intervals assume you rate each review "Good". Rating "Easy" increases the interval multiplier; rating "Hard" or "Again" decreases it and increases frequency.
In steady state — after you have been reviewing a deck for several months — most cards appear only once per month or less. The initial investment in frequent reviews produces a system that eventually runs on near-autopilot.
What happens when you skip reviews
Missing one review session rarely causes catastrophic forgetting. The memory will be somewhat harder to retrieve, but the forgetting curve's current level is usually still above zero. The danger is cascade skipping: missing one day, then another, then another.
After three or four consecutive skipped days on material you have reviewed only once or twice, forgetting may be significant enough to require near-complete re-study. The discipline required is not intensity — it is consistency. A 10-minute session on a difficult day is worth far more than skipping and catching up.
The review session itself
A review session is not re-reading. The structure for maximum benefit:
- Cover the answer (card back, or close your notes)
- Attempt retrieval — produce the answer entirely from memory
- Uncover the answer — check what you got right and wrong
- Rate your recall — in Anki, use Again/Hard/Good/Easy; in a manual system, pass/fail
- Move on — do not dwell on correct answers; focus extra time on failed items
The review session for a 200-card deck at steady state takes 15–25 minutes. New cards that have been reviewed only once or twice will take longer per card — this is expected, and necessary.
Build a deck and try it
The fastest way to understand spaced repetition scheduling is to experience it. Build a small deck of 20–30 cards on something you are currently studying and follow the schedule above for 30 days.
Use the WarpRead Flashcard Tool to create your deck in-browser, enter focus mode for daily reviews, and export as HTML for offline sessions. The free Spaced Repetition course covers the full research behind optimal scheduling in Lesson 2, with worked examples from Cepeda's meta-analysis.
References
- Cepeda, N.J., et al. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380.
- Ebbinghaus, H. (1885). Über das Gedächtnis. Duncker & Humblot.
- Wozniak, P.A., & Gorzelanczyk, E.J. (1994). Optimization of repetition spacing in the practice of learning. Acta Neurobiologiae Experimentalis, 54(1), 59–62.
- Dunlosky, J., et al. (2013). Improving students' learning with effective study techniques. Psychological Science in the Public Interest, 14(1), 4–58.
Topics
Build your spaced repetition deck
Create atomic flashcards in-browser, import from an AI-generated .txt file, and enter Focus Mode for random-order paper-card review. Export as a standalone HTML for offline sessions. Free, no account.
More on Spaced Repetition & Flashcards