Spacing and retrieval practice
Claim assessed: Spreading study over several sessions and testing yourself lead to better retention than massed rereading.
Published , updated · how we work
In short
Two of the best-documented findings in the psychology of learning go together:
- Spacing: you remember more if you split the same study time into several sessions, spread apart, than if you learn everything at once.
- Retrieval practice: you remember more if you try to recall the material without looking at it than if you reread it.
Both share the same trap: in the moment, the weaker option feels better. Rereading and cramming give a feeling of knowing, but the memory lasts less.
What it says
Testing beats rereading, but not right away. Roediger and Karpicke (2006) gave students short texts to read, about the sun or about sea otters. After a first reading, some of them reread the text, and the others wrote down everything they remembered from it, without being given any correct answers. The final test came after 5 minutes, 2 days or a week.
In the first experiment, after 5 minutes, those who had reread remembered slightly more: 81% versus 75%. After 2 days, the situation reversed: 68% for those who had been tested, versus 54%. After a week: 56% versus 42%. After a week, the tested students remembered about as much as those who had reread remembered after only 2 days.
The second experiment went further, with four 5-minute sessions:
One important detail: those who had read the text four times were the most confident that they would remember it a week later. They were exactly the ones who remembered the least after a week.
The result is not isolated. The meta-analysis by Adesope, Trevisan and Sundararajan (2017) gathered 272 comparisons from 118 articles, with 15,427 participants. Practice tests produced better results than rereading and than every other option compared: the average effect was g = 0.61, and against rereading, g = 0.51. The g index shows how large the difference between groups is; by Cohen’s conventions, 0.5 means a medium effect. The authors checked for selective publication only with an old method that doesn’t correct the effect.
A newer meta-analysis, by Yang, Luo, Vadillo, Yu and Shanks (2021), was limited to classroom studies: 222 studies, with 48,478 pupils and students. Testing improved academic results by g = 0.50. Corrected for unpublished studies with several methods, the effect stayed almost the same, between 0.43 and 0.48. Against rereading the advantage was smaller, g = 0.33, and against other elaborative strategies, such as concept mapping, almost nil (g = 0.10).
Spacing: it depends how long you want to remember. Cepeda and colleagues (2006) gathered 839 measurements from 317 experiments, published in 184 articles. They found that the optimal gap between sessions grows the longer you want to remember the material.
A later study by the same team (Cepeda et al., 2008) measured this directly, with more than 1,350 participants. They learned a set of facts, got a review after a gap of up to 3.5 months, and took a final test up to a year after that. The optimal gap was about 20–40% of the time until the test when the test was a week away, and 5–10% when it was a year away. Our own arithmetic from these proportions: for a test in a week, the ideal review comes after about 1.5–3 days. For something you want to remember for a year, after about 2.5–5 weeks.
The overall assessment. Dunlosky and colleagues (2013) evaluated ten learning techniques used by students. Only two received a high-utility rating: practice testing and distributed practice, because they help people of different ages and abilities, on many kinds of assessment, including in school. Rereading and highlighting, though among the most used, received low utility.
Example
An example built for this text: you have to learn 40 new terms for an exam in two weeks. The usual approach is to read the list a few times the night before the exam. The approach the studies support: today you learn them and cover the definitions to try to recall them. A few days later, you again try to recall each definition before looking. You repeat once more a few days before the exam. The total time can be the same. The difference is how and when you use it.
How to apply it
The steps below are a practical approach we propose, based on the studies cited.
- Close the book and try to recall. Write down or say out loud everything you know about the topic, then check. In Roediger and Karpicke, testing helped even without the correct answers, but checking keeps you from remembering mistakes.
- Split your studying into several sessions. The same number of hours spread over several days gives better results than one long session.
- Match the gap to the deadline. The longer you need to remember, the longer the gaps between reviews can be. As a starting point, use the proportions from Cepeda et al. (2008).
- Don’t trust the feeling. Rereading feels more effective. In Kornell and Bjork’s (2008) first experiment, 78% of participants learned better from spaced examples, yet 78% rated the massed version as equally good or better, even after the test that had shown them the opposite.
- Use flashcards or spaced-repetition apps. They combine both ideas: they ask you to recall the answer and schedule reviews at growing intervals. The specific apps were not tested in the studies here.
Limits and nuances
- Most studies use simple materials. Words, word pairs, facts or short texts, tested by recall. Dunlosky et al. (2013) note, though, that both techniques have also been tested in real school settings.
- In the very short term, rereading can win. If the test is a few minutes away, as with an exam given right after studying, the advantage of testing disappears or reverses.
- The size of the effect varies. The meta-analyses show that the benefit depends on the type of test, the material and the participants. It is not a fixed number. In the classroom, the advantage over rereading is rather small, and against other active strategies it almost disappears.
- For spacing we have no corrected estimate. Cepeda et al. (2006) did not correct the effect for unpublished studies. They argue that the effect is too large to be explained by studies left unpublished.
- The proportions for the optimal gap are a guide. They come from a single large study in which participants learned a set of simple facts, and they change with the time until the test.
- Spacing also helped concept learning, in one study. Kornell and Bjork expected spacing to make it harder to see what a category’s examples have in common. The result was the opposite: spacing helped there too, in a task about recognizing painters’ styles.
- What we read. We had the full text of all the sources and checked the figures cited here against it: Roediger and Karpicke (2006), from a copy on the East Carolina University website; Kornell and Bjork (2008), from the Bjork lab website; Cepeda et al. (2006) and Cepeda et al. (2008), from the University of California repository, the second in its accepted version; Dunlosky et al. (2013), Adesope et al. (2017) and Yang et al. (2021), from copies on gwern.net. The proportions for the optimal gap are those in the abstract of the published version of Cepeda et al. (2008). Of the three long works, Dunlosky, Adesope and Yang, we read the abstracts and the sections on the main results, not every analysis.
- Revised on 7 October 2026. We added the meta-analysis by Yang et al. (2021), with the effect corrected for unpublished studies, and the main figures from Adesope et al. (2017). We made the Kornell and Bjork result more precise and added this list.
Sources
- Henry L. Roediger III, Jeffrey D. Karpicke (2006). Test-enhanced learning: Taking memory tests improves long-term retention
- Nicholas J. Cepeda, Harold Pashler, Edward Vul, John T. Wixted, Doug Rohrer (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis
- Nicholas J. Cepeda, Edward Vul, Doug Rohrer, John T. Wixted, Harold Pashler (2008). Spacing effects in learning: A temporal ridgeline of optimal retention
- Nate Kornell, Robert A. Bjork (2008). Learning concepts and categories: Is spacing the "enemy of induction"?
- John Dunlosky, Katherine A. Rawson, Elizabeth J. Marsh, Mitchell J. Nathan, Daniel T. Willingham (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology
- Olusola O. Adesope, Dominic A. Trevisan, Narayankripa Sundararajan (2017). Rethinking the use of tests: A meta-analysis of practice testing
- Chunliang Yang, Liang Luo, Miguel A. Vadillo, Rongjun Yu, David R. Shanks (2021). Testing (quizzing) boosts classroom learning: A systematic and meta-analytic review
See also: Implementation intentions, Interleaving, Learning styles, Serial position effect, Zeigarnik effect