Spaced repetition: what the spacing effect research says and how to schedule reviews

Spaced repetition and the spacing effect

Spaced repetition means reviewing the same material in several sessions spread over days or weeks, instead of in one long block. The memory benefit it produces is called the spacing effect (researchers also say distributed practice). Its opposite is massed practice, better known as cramming.

The spacing effect is one of the most replicated findings in learning research. A major review rated distributed practice as one of only two "high utility" study techniques, alongside practice testing (Dunlosky et al., 2013). A later meta-analysis of the same ten techniques, covering 242 studies, found distributed practice and practice testing were the two most effective (Donoghue & Hattie, 2021).

The useful question is not whether to space, but how far apart. The research has a fairly precise answer.

What the research shows

The best gap depends on when the test is

Cepeda et al., 2006 pooled 839 assessments from 317 experiments in 184 articles on verbal learning. Their central finding: the gap between study sessions that gives the best retention increases as the retention interval (the time until the test) increases. A gap that is perfect for a test next week is too short for a test in three months.

Cepeda et al., 2008 then tested this over realistic timescales. More than 1,350 people learned a set of facts, reviewed them after a gap of up to 3.5 months, and took a final test up to a year later. Two results matter for anyone with an exam date:

• At any test delay, lengthening the gap first improved final test performance and then gradually reduced it. There is a sweet spot.
• Measured as a share of the time until the test, the best gap fell from about 20 to 40% of a 1-week test delay to about 5 to 10% of a 1-year test delay.

Cepeda et al., 2009, with test delays of up to 6 months, found an optimal gap improved final recall by up to 150%. Accuracy rose sharply as the gap grew and then declined only gradually, so a gap that is a bit too long costs less than one that is too short. The authors also stress that cumulative reviews matter.

It works in real classrooms, not only labs

Most spacing studies use word lists and facts in a lab. A 2025 meta-analysis of applied classroom studies (22 reports, 31 effect sizes, more than 3,000 learners) found a moderate effect in favour of distributed over massed practice, d = 0.54 (Mawson et al., 2025). In plain English, d = 0.54 means the average student who spaced their practice scored about half a standard deviation higher than the average student who crammed. Effects were larger with longer retention intervals and at higher education levels, which describes most certification candidates.

In a precalculus course for engineering students, Lyle et al., 2020 found that within-semester retention benefited from both more practice and more spacing, but retention four weeks later, across semesters, benefited only from more spacing. More practice did not substitute for spreading it out.

Why spacing works and why cramming feels better

Researchers still debate the exact mechanism. One common account is that when some forgetting has happened, the next review takes more effort to retrieve, and that effort strengthens the memory more. This fits the idea of desirable difficulties: conditions that feel harder while studying but improve later performance (de Bruin et al., 2023).

Cramming feels better because everything is fresh. On a test the same evening, massed study can look fine. The cost shows up days later, when cramming has left little behind. See the forgetting curve for how fast unrehearsed material fades, and how each review flattens the curve.

Spacing combines naturally with active recall. The review sessions should be self-tests, not rereading. A flashcard box system such as the Leitner system is one way to automate both: cards you answer correctly come back after longer gaps, cards you miss come back soon.

A spaced repetition schedule for your exam date

The table below is our application of the Cepeda findings to typical certification timelines. It is not a figure from the papers. The papers give the optimal gap as a share of the test delay for single review sessions; we have rounded those shares into review days and added cumulative reviews, as Cepeda et al. (2009) recommend. For delays between 1 week and 1 year, the papers do not give an exact share, so our middle rows are an interpolation.

How to read it: find the row for how far away the exam is for the topic you are learning today. A topic learned in week 1 of a 12-week plan follows the 12-week row; a topic learned in week 11 follows the 1-week row.

Three rules hold in every row:

• Each review is a self-test, then a check of the answers.
• When in doubt, wait a day longer, not shorter. In Cepeda's data, too-short gaps hurt more than slightly long ones.
• Reviews are cumulative. Old topics keep coming back in mixed sets until the exam.

Tools: apps, boxes and a calendar

You do not need special software to space your study. Three approaches work, and they suit different kinds of exam content:

• Spaced repetition flashcard apps. They schedule each card individually, showing cards you miss more often and cards you know less often. Good for high-volume facts: port numbers, acronyms, command syntax, formula definitions. Keep cards small, one fact per card, and write them yourself from the official exam objectives.
• A paper box system. The Leitner system does the same thing with physical cards moving between boxes. It is slower, but writing cards by hand is itself a round of processing.
• A calendar of topic reviews. For scenario-based exams such as the PMP or CISSP, cards cover only part of what you need. Schedule whole-topic reviews instead: a set of practice questions and a blank-page summary for each domain, placed on the days the table above suggests.

Whichever tool you use, check two things each week: that old topics are still coming back, and that your review time is not crowding out practice questions.

Common mistakes

• Spacing rereading instead of retrieval. A spaced reread is better than a crammed one, but a spaced self-test is better still.
• Reviewing too soon. Reviewing the same evening feels productive but adds little; let some forgetting happen.
• Stopping reviews once a topic feels known. Familiarity is not the same as recall. Keep the topic in rotation.
• Planning only new material. Book review sessions in your calendar the same way you book new chapters. A study schedule with review slots makes this automatic.
• Expecting it to feel efficient. Spaced practice often feels slower. Judge it by your practice test scores a week later.

Limits of the evidence

Most studies, including the large Cepeda experiments, test factual or verbal material: word pairs, trivia, vocabulary. The classroom meta-analysis by Mawson et al. is reassuring, but it covers 31 effect sizes, not hundreds. And Donoghue and Hattie caution that most studies in their meta-analysis measured surface or factual outcomes. For complex skills such as subnetting or troubleshooting, spacing is still sensible, but the optimal gaps are less precisely known.

How to use this for a certification exam

Example: CompTIA Security+ with the exam 8 weeks away and five study days per week.

1. Day 1. Take our free Security+ practice test cold to see your weakest domains.
2. Each study day: 20 minutes of reviews before new material. Reviews come first so they never get skipped.
3. Review timing. For topics learned in weeks 1 to 4, review after about 5 to 7 days, then 2 weeks later, then in the final week. For topics learned in weeks 5 to 7, review after 3 to 4 days and again in the final week. Topics learned in the last week get one review 2 to 3 days later.
4. What a review is. 10 to 15 questions per topic, answered from memory, then checked. Missed items come back after 1 to 2 days.
5. Every second week, a mixed set. 30 to 40 questions across all domains, plus a retake of our free practice test. Track domain scores against your day-1 baseline.
6. Final week. No new topics. Daily mixed review sets of 20 to 30 questions, with your weakest domains over-represented.

For other exams, use the CCNA practice test or the test for your certification in the same way. For a complete plan from 8 weeks out, see how to study for exams.