
There is no scientifically established ideal number of flashcards to study per day. Research suggests that how you study matters more than how many cards you get through: pile size in a sitting, how many times you recall a card, and how long you wait before it returns.
The figures below come from free peer-reviewed papers. A few sums and one sample routine are labelled as ours. Superscripts jump to References.
New cards vs daily reviews
"How many flashcards a day?" usually mixes two jobs. New cards are material you add today. Due cards are ones that return from earlier days. Adding 30 new cards does not mean studying only 30: old cards come back on a schedule.
| Example sitting | |
|---|---|
| New cards | 15 |
| Due cards | 35 |
| Total to study | 50 |
Illustrative only, not a research target. Harder cards and more misses take longer. If due work already fills the time you have, add fewer new ones.
With a spaced repetition app like Leitiq, you do not pick the due pile by hand each day. The queue follows your progress so new material and reviews stay in balance.
What the studies found
Small stacks feel better than they work
Kornell compared the same 40 GRE-style pairs (for example "effulgent: brilliant") in a stack of 20 versus four stacks of five. Every card was studied the same number of times.1
| Experiment | Large stack (20) | Small stacks (4×5) |
|---|---|---|
| 1: test the next day | 49% | 36% |
| 2: four study days, test on day 5 | 54% | 21% |
| 3: plus a review before the test | 65% | 34% |
Kornell combined the three experiments (20, 25 and 25 participants) and reported that spacing helped 63 of those 70 (90%). Time per card did not differ. After session 1, 72% of those with a preference still thought the small stacks had worked better:1
Massed study is seductive, and it can appear to be more effective than spaced study even when spaced study is the more effective strategy.
"To be efficient, flashcards should be studied in relatively large stacks across multiple days."1 The 20-40 items in these lab sittings are a useful reference, not a universal daily target. The cards were light vocabulary pairs. None of the studies here tested 100.1234
Count recalls, not first-time wins
Many learners drop a card after one correct answer. With English-Swahili pairs, Kornell and Bjork found that being allowed to drop cards "had small but consistently negative effects on learning." Students usually dropped after a single correct response:2
| Correct recalls before drop | Later recall |
|---|---|
| 0 | 19% |
| 1 | 62% |
| 2 | 88% |
Forcing a drop after one correct recall did worse than never dropping. They called that "a maladaptive strategy" and said "items, if dropped, should be returned to later."2
Karpicke and Roediger then separated re-reading from testing on 40 Swahili-English pairs (mashua-boat). After the first correct recall, pairs that stayed in the tests were recalled about 80% a week later. Pairs dropped from tests were at 36% (kept in study) and 33% (dropped from both), even when students kept re-reading. "Repeated study after one successful recall did not produce any measurable learning a week later."3
One study of 533 students (cited here from its free abstract) found that practicing to about three correct recalls in a first session, then relearning a few times at spaced intervals, improved long-term retention with little extra practice cost. That is a useful starting point, not a rule for every card or system.5
How long to wait depends on how long you need the memory
Cepeda and colleagues taught more than 1,350 adults 32 obscure facts, with a gap of up to 3.5 months between two learning sessions and a final test up to a year later.4
| When you need the fact | Best gap between the two learning sessions |
|---|---|
| About 1 week after the review | About 3 days (~20-40% of that delay) |
| About 1 year after the review | About 27 days (~5-10% of that delay) |
The best gap raised final recall by 64% versus reviewing on the same day. "if you want to know the optimal distribution of your study time, you need to decide how long you wish to remember something."4 These are experimental averages under specific retention goals. Do not copy the 27-day figure into an ordinary daily flashcard schedule.
Murre and Dros's one-subject Ebbinghaus replication shows the same shape: the longer the wait, the closer relearning gets to starting over ("similar to Ebbinghaus' original data").6
In a microbiology course, instructor Anki decks reviewed on four days a week did not change overall exams significantly, but Exam 1 and confidence rose, and flashcard time share went from 3.25% to 29.2% of 3-4 study hours a week (about 53-70 minutes, our arithmetic). The authors warn the software "can be easily clicked through" and that spaced repetition resources "yield ineffective results in retention when used to cram content."7 Kornell's lab matches that: eight trials crammed before the test → 34% recalled; the same eight spread across days → 54%.1
Tabibian and colleagues, on observational Duolingo data, treat reviews as a cost against recall: weaker cards should return sooner. That supports adaptive scheduling. It is not an experiment that proves one daily card count.8
How to find your own number
Our synthesis, not a finding of the papers:
- Clear due cards before stacking on new ones.
- Prefer one finishable pile over tiny massed stacks; treat 20-40 light items as a lab reference, then adjust for difficulty and time.1
- Do not drop a card after a single correct recall; a few solid retrievals plus later returns beat one lucky hit.235
- Set gaps from how long you need the memory, not from a fixed "27 days."4
- If due work outgrows your time, cut new cards first.8
- Spread study across days; cramming underperforms spacing in the lab and the classroom data.17
In Kornell's Experiment 3, mean total study time per card was 41.47 seconds in the large-stack condition. That figure already sums cue and answer time across all ten trials, not seconds per flip. For 20 light cards that is about 14 minutes across the five sessions (41.47 × 20 ÷ 60), our arithmetic. Harder material takes longer.1
FAQ
How many new flashcards per day?
No paper here sets a new-card quota. Start small enough that you can still finish today's due pile, then raise only if reviews stay manageable.
How long should you study flashcards?
Use a time cap you can keep most days. Light lab cards can move in a few minutes; dense cards will not.
How many flashcards should you review?
Review what is due. That count rises and falls with how many new cards you added in prior weeks.
Is studying 100 flashcards a day too much?
None of these studies tested 100. A huge sitting is usually too many new cards, an overdue backlog, or both. Shrink new intake until the queue fits your time.
How often should you review flashcards?
Often enough that cards return before they are gone, with wider gaps when you need longer retention. A scheduler does that bookkeeping for you.
What these papers cannot tell you
Most evidence is short lab work: word pairs, trivia, nonsense syllables. Murre and Dros used one subject.6 The classroom study was not randomized and could not verify how cards were used.7 The Duolingo analysis is observational.8 Rawson and Dunlosky is cited from its free abstract only.5 None of the papers tests a daily count.
Build a daily flashcard habit with Leitiq
You do not need hundreds of cards every day. Start building a daily habit with Leitiq: create your first deck, study what is due, and let spaced repetition organize your reviews. It is free to get started: create an account.
How the app schedules those returns: How Session Leitner works in Leitiq and How FSRS interval repetition works in Leitiq.
References
Kornell N. Optimising learning using flashcards: Spacing is more effective than cramming. Applied Cognitive Psychology. 2009;23:1297-1317. doi:10.1002/acp.1537. Free PDF. ↩ ↩ ↩ ↩ ↩ ↩ ↩ ↩
Kornell N, Bjork RA. Optimizing self-regulated study: The benefits and costs of dropping flashcards. Memory. 2008;16:125-136. doi:10.1080/09658210701763899. Free PDF. ↩ ↩ ↩ ↩
Karpicke JD, Roediger HL III. The critical importance of retrieval for learning. Science. 2008;319(5865):966-968. doi:10.1126/science.1152408. Free PDF. ↩ ↩ ↩
Cepeda NJ, Vul E, Rohrer D, Wixted JT, Pashler H. Spacing effects in learning: A temporal ridgeline of optimal retention. Psychological Science. 2008;19(11):1095-1102. doi:10.1111/j.1467-9280.2008.02209.x. Free PDF. ↩ ↩ ↩ ↩
Rawson KA, Dunlosky J. Optimizing schedules of retrieval practice for durable and efficient learning: How much is enough? Journal of Experimental Psychology: General. 2011;140(3):283-302. doi:10.1037/a0023956. Free abstract. ↩ ↩ ↩
Murre JMJ, Dros J. Replication and analysis of Ebbinghaus' forgetting curve. PLoS ONE. 2015;10(7):e0120644. doi:10.1371/journal.pone.0120644. Free full text. ↩ ↩
Ogunjobi F, Alexander SM, Cramer L. Usage of spaced repetition flashcards to improve educational outcomes in microbiology. Cureus. 2024;16(10):e70994. doi:10.7759/cureus.70994. Free full text. ↩ ↩ ↩
Tabibian B, Upadhyay U, De A, Zarezade A, Schölkopf B, Gomez-Rodriguez M. Enhancing human learning via spaced repetition optimization. Proceedings of the National Academy of Sciences. 2019;116(10):3988-3993. doi:10.1073/pnas.1815156116. Free full text. ↩ ↩ ↩