Spaced repetition is one of the most recommended study techniques in medicine, and one of the most overclaimed. So it is worth asking what the evidence actually shows, and in 2026 there is a strong answer: a systematic review and meta-analysis of 21,415 learners found a large effect in favour of spaced repetition over standard study. That is a genuine, well-evidenced finding, but it comes with real limitations that most enthusiastic blog posts skip. Here is the honest picture.
Key takeaways
- A 2026 meta-analysis of 21,415 learners found spaced repetition beat standard study with an SMD of 0.78.
- That is a large effect by the standards of educational research, where effects are usually smaller.
- The evidence has real limits: mixed interventions and mostly short-term, objective-test outcomes.
- Supporting reviews and the classic retrieval-practice literature point the same way.
- A good spacing system re-surfaces what you get wrong at widening intervals, not everything equally.
The claim everyone makes, and the evidence it rested on
For years, the case for spaced repetition leaned on the same two pillars: Ebbinghaus's nineteenth-century forgetting curve, and the retrieval-practice experiments showing that testing yourself beats rereading. Those are real and important, but they are old and narrow foundations for a claim that gets repeated with total confidence across medical education. What was missing was a rigorous synthesis of whether spaced repetition actually improves performance in medical settings specifically, rather than in psychology laboratories.
The 2026 anchor
That synthesis now exists. Maye and colleagues, publishing in The Clinical Teacher in 2026, conducted a PRISMA-compliant systematic review and meta-analysis, screening 542 records, including 14 studies in the review and 13 in the meta-analysis, covering 21,415 learners. The headline result was an overall significant effect in favour of spaced repetition over standard study techniques, with a standardised mean difference of 0.78 (95% confidence interval 0.56 to 0.99, p < 0.0001). This is the strongest single anchor to cite for spaced repetition in medical education, and it is recent and specific to the field.
What an SMD of 0.78 actually means
A standardised mean difference expresses the gap between two groups in units of variability, and 0.78 is a large effect by the norms of educational research, where interventions typically produce much smaller effects. Intuitively, it means the average learner using spaced repetition outperformed roughly three-quarters of those using standard study. That is a substantial edge, and the tight confidence interval, well above zero, means the effect is reliable rather than a statistical fluke. It is not a magic multiplier, but it is a real and meaningful advantage.
The honest limitations
Now the parts most pages leave out. The interventions studied were heterogeneous, ranging from faculty-created and third-party flashcards to multiple-choice questions delivered by email or within continuing medical education frameworks, to spaced classroom quizzes, so "spaced repetition" here is a family of methods, not one tool. Most outcomes were short-term performance on objective tests, so the evidence for long-term retention and for continuing medical education is thinner. And the authors are explicit that the optimal design and delivery of spaced repetition, the ideal intervals, formats, and dosing, remain open questions. The effect is real; the recipe is not settled.
Supporting evidence
The meta-analysis does not stand alone. A separate systematic review of spaced digital education for health professionals found spaced online education superior to massed online education for knowledge and for clinical behaviour change, with smaller but positive effect sizes, and larger effects for practical skills. Beneath both sits the well-established retrieval-practice literature showing that repeated testing produces markedly better long-term retention than repeated study. The direction of evidence is consistent across syntheses: spacing plus retrieval works.
What this means for how you study
The translation is specific. A good spacing system does not re-show you everything at fixed intervals; it re-surfaces what you got wrong, and what you are about to forget, at widening intervals, so your time concentrates on your genuine weak points. That is the mechanism behind the effect: not more exposure, but better-timed retrieval of the right material. iatroX builds spacing and retrieval into its engine so your revision targets your misses over time rather than re-covering what you already know, with free sample questions to try at iatroX. We will not claim it is proven to raise your score, because that is not what the evidence says; the evidence supports the method, and the method is what iatroX implements.
Frequently asked questions
Does spaced repetition work for medical exams? Yes, on the current evidence. A 2026 meta-analysis of 21,415 learners found a large effect in its favour over standard study, with a standardised mean difference of 0.78, and supporting reviews agree.
How big is the effect? An SMD of 0.78 is a large effect by educational research standards. Intuitively, the average spaced-repetition learner outperformed about three-quarters of those using standard study, with a reliable, tight confidence interval.
What are the limits of the evidence? The studied interventions were varied, most outcomes were short-term objective tests, and the optimal intervals and formats are still unsettled. The method works, but the ideal design is not yet defined.
Is Anki the only way to do spaced repetition? No. The evidence covers a family of methods, including flashcards, spaced multiple-choice questions, and spaced quizzes. What matters is spaced, repeated retrieval of the right material, not one specific tool.
Does spaced repetition help long-term retention? The retrieval-practice literature supports better long-term retention, but the meta-analytic evidence is strongest for shorter-term objective-test performance, with longer-term effects less thoroughly studied.