The Spacing Effect — Spread It Out and It Lasts

The Same 3 Hours, Two Plans

There is an exam in a month. The total time available to study is 3 hours. Minho sits down on Wednesday evening and finishes it all in one 3-hour block. Sua studies the same material for one hour each on Monday, Tuesday, and Wednesday. Both finish studying on Wednesday, and both spend the same 3 hours.

On Wednesday night, Minho feels accomplished. The material he just looked at is sharp in his head, and if he opens the book again, it reads smoothly. When Sua opened her book on Tuesday, what she had seen on Monday was already hazy, so she felt a little frustrated. Just looking at that night alone, it feels like Minho knows the material better.

Then, who will remember more on the exam a month from now? Starting from that question, this article covers how to space out your study sessions and how to mix up your problem-solving.

Let's carry one picture all the way through. Suppose we need to pour three cups of water into a flowerpot. If we pour all three cups at once, the soil looks thoroughly soaked that day, but a lot of the water escapes into the saucer. If we give one cup a day for three days, the water seeps into the soil. Minho poured all three cups at once, and Sua gave them one cup at a time.

This flowerpot and the plans of Minho and Sua are examples created by this article to help understanding. The flowerpot is merely a picture to help imagine the result. This does not mean that memories overflow like water in the head, and regarding why spacing things out makes them last longer, there are several explanations among researchers (PMID 35950522, PMID 41508718).

Spreading It Out Leaves More Behind

The two plans have names. Studying the same amount of time all at once is called massed practice, and doing it multiple times with spaces in between is called spaced practice. Minho's Wednesday evening is massed practice, and Sua's Monday, Tuesday, and Wednesday is spaced practice.

If you studied the same material for the same total amount of time, the part you did spaced out usually remembers more on a later test. The phenomenon where studying spaced out over intervals leaves more behind for a long time is called the spacing effect. An analysis compiling experiments that memorized and recalled words or facts included over 180 papers (DOI: 10.1037/0033-2909.132.3.354).

Outside the laboratory, the direction is the same. In an analysis of 22 studies conducted in classrooms using materials and periods suited to curricula, gathering over 3,000 students, studying spread out outperformed studying crammed all at once by a medium effect size (PMID 40564553).

A comprehensive evaluation comparing ten learning methods placed spaced practice in the highest tier alongside practice testing. The reason is that the effect appeared across students of different ages and abilities, on various types of tests, and even in real classrooms (PMID 26173288).

Sua's plan aligns with the direction of these studies. If the test is a month away, Sua is very likely to remember more overall.

If you spend the same amount of time, spreading it out rather than cramming it all at once usually leaves more behind until later.

That Night's Sense of Fulfillment

Minho's sense of fulfillment on Wednesday night also has a reason. When you cram, the material you just looked at is still in your head, so things actually go well on the spot. It is for the same reason that cramming the night before an exam feels like it is working well.

Methods that help things stick for a long time, like spacing out, tend to actually make studying slower when you are first learning the material. Instead, they help with keeping things for a long time and using what you learned on new problems (PMID 41508718, PMID 35950522). The frustration Suah felt on Tuesday is this slowness.

There is also a clue that the difference looks larger the further away the exam is. In the classroom study analysis mentioned earlier, studies with a longer period from finishing the study to the final exam generally showed greater benefits for spacing out (PMID 40564553). However, the analysis notes that because the number of studies was small, that trend could not be formally verified with numbers. To put it in terms of flowerpots, it means that just by looking at soil that seems wet that evening, you cannot tell which side's soil will still be holding moisture a few days later.

What cramming relies on is the smoothness of that night. If it is material you need to carry with you until next month or next semester, the spaced-out side generally leaves behind more. How much of a difference there is between the two methods when the exam is very close is hard to say based on the studies this article looked at.

The feeling that studying you crammed went well on the spot is not a sign that it will stay with you for a long time.

How Long to Leave Between

Suh-ah spaced them out by a day. Would it be better to leave a week instead of a day?

The answer depends on how long you want to remember. The interval that left the most behind moved along with the time until the final test, and the longer you needed to remember, the longer the best interval became (DOI: 10.1037/0033-2909.132.3.354). In the records of just over 10,000 people who received job training at work, the shape of the laboratory—where the best interval gets longer the longer you need to remember—appeared just the same (PMID 30623389). The interval for next week's exam and the interval for material to remember until next year are different.

We also frequently hear that intervals should gradually widen, such as 1 day, 3 days, and 7 days. In one laboratory experiment, widening the overall interval helped a lot in retaining information for a long time, but whether the intervals gradually widened, stayed the same, or gradually narrowed, that shape itself was not superior (DOI: 10.1037/a0023436). Because this is the result of a single laboratory experiment, it cannot be directly applied to all studying.

Fixed formulas like "so many days later, so many times" do not exist in the studies this article relies on. What you should take away is the direction.

The longer you need to remember, the wider you set the interval. Exactly how much is just right depends on your target duration.

Interleaving

Let's think about a math homework assignment. If you learned the Pythagorean theorem today, the homework is usually twelve problems about the Pythagorean theorem. Even before reading the problems, you already know what method to use (DOI: 10.1037/edu0000001).

Solving all problems of one kind before moving on to the next kind is called blocked practice. Interleaving is practice where you mix several kinds of problems together so that the same method is not needed consecutively. In math, it is homework designed so that two back-to-back problems do not require the same method (DOI: 10.1037/edu0000367). In interleaved practice, you have to look at a problem and choose the method yourself. You are in the same situation as when you encounter problems on a comprehensive exam or in next year's class (DOI: 10.1037/edu0000001).

When you solve things mixed together, you make more mistakes while practicing. There is an experiment where children were asked to solve four kinds of math problems either mixed or blocked, with the time gaps between the same kinds kept identical. The mixed group had worse scores during practice, but on a test one day later, their score was twice as high (DOI: 10.1002/acp.1598). Because the time gaps were kept equal, this difference can be seen as coming from the mixing itself, not from spacing.

When looking closely at the wrong answers, the children in the mixed group were better at pairing each problem with the correct method. The benefit of mixing seems to come from the ability to choose "what method goes with this problem" (DOI: 10.1002/acp.1598).

Results came out in actual classrooms, too. Fifty-four seventh-grade (roughly 12-13 years old) math classes received either mixed homework or blocked homework for four months, and at the end, both groups did mixed review homework. On an unannounced test one month later, the mixed group scored 61% and the blocked group scored 38% correct (DOI: 10.1037/edu0000367).

The researchers of this classroom study noted that important limitations remain to be examined, yet concluded that interleaved math practice is effective and worth trying in the classroom (DOI: 10.1037/edu0000367). However, because mixed homework also scatters problems of the same kind across multiple days, a portion of this difference may also come from spacing (DOI: 10.1002/acp.1598).

When things are mixed, the differences between categories become visible side by side. When learning paintings by multiple artists, just leaving time between paintings by a single artist was not helpful, and the group that mixed paintings by multiple artists or placed them side by side was better at guessing the artist of a previously unseen painting (DOI: 10.1002/acp.1801). The authors suggest that the effect comes not from the time gaps, but from the contrast between the styles of the artists.

Solving mixed up makes you make more mistakes on the spot, but later on, it seems to help you better choose "what method goes with this problem".

Mixing Doesn't Always Win

Mixing comes with more conditions attached than block studying. In an analysis combining 59 studies, the benefit of mixing was medium-sized overall. It varied by material. It was largest for paintings, small for math problems, and not distinct for explanatory texts and taste discrimination. For word lists, studying them grouped together was actually better (PMID 31556629).

A pattern also appeared regarding which materials it works well for. The benefits were greater the more similar the categories were to each other, the more different the examples within a single category were from each other, and the more complex the materials were. The authors concluded that mixing is useful for studying categories by looking at multiple examples, but that it should be used by considering the type of material and the situation, and with caution especially for explanatory texts and words (PMID 31556629).

What you know when you start also matters. In a field experiment where 176 high school students learned a difficult physics unit, students with more background knowledge did relatively better with mixing, while those with less knowledge seemed to do better with block solving (DOI: 10.31234/osf.io/m46vf_v1). This was the result of a test right after studying; on a test 8 weeks later, this difference no longer appeared. This is the result of a preprint, a manuscript made public before peer review.

There are also classrooms where the effect of mixing did not appear again. In two large classroom experiments teaching mathematically defined concepts with clear meanings, such as angle relationships, no difference was caught between mixing and blocking (DOI: 10.5951/jresematheduc-2024-0055). The same report points out that a significant portion of the evidence recommending mixing has lumped together the effects of mixing and block studying.

The earlier comprehensive review also gave mixing a medium grade in 2013, one step lower than block studying. The reason was that research systematically examining the benefits of mixing had only just begun (PMID 26173288).

Mixing works best when you need to distinguish similar categories from one another, and depending on the materials and background knowledge, the benefits can disappear or even reverse.

Why Harder Is Better

Spaced practice and interleaved practice share something in common. Both feel harder while you are doing them. Sua on Tuesday had to grab at fading memories, and the student given the mixed-up homework had to figure out the method each time.

When conditions that look harder and slower during studying are actually advantageous for long-term memory, those conditions are called desirable difficulties. Well-supported examples of desirable difficulties across several fields include testing during learning, interleaved practice, and spaced practice (PMID 41508718).

The problem is that students read this difficulty as a bad sign. In two studies where college students were asked to design a hypothetical math study schedule or choose among five schedules, most wrongly judged the schedules with the least spacing and interleaving as the most effective (DOI: 10.1037/xap0000391). Schedules with spacing and interleaving were seen as harder, less enjoyable, and less common.

The same pattern appeared in experiments that had people directly experience interleaved versus blocked problem-solving, or retrieval practice versus restudy. The methods they felt were harder were rated as less effective, and thus they chose them less for their next study session (PMID 31470194). Among those, in an experiment comparing retrieval practice and restudy, contrary to the students' judgments, choosing the harder side was linked to longer retention. The authors explain this with the hypothesis that it happens because people misread the feeling of difficulty as a sign that they are not learning.

We must also be careful of the illusion in the opposite direction. The same review warns that feeling harder should not be taken right away to mean learning more, and that we must clearly know and use what is a desirable difficulty and what is not (PMID 41508718).

The feeling of difficulty may not be a bad sign. But not all difficulty is good either.

Two Weeks Before the Exam, One Page on the Calendar

Sua has an exam two weeks from now. The material covers three units. Let us make a calendar using only what we have seen so far.

First, we space the material out. Instead of cramming three units into a single weekend, we spread them across multiple days over two weeks. If the exam is two weeks away, a shorter gap is more appropriate than an interval meant for several months later, because the best gap changes along with the target period (DOI: 10.1037/0033-2909.132.3.354). If the content is something you want to keep even after the exam passes, you leave a longer gap and review it one more time further out.

On the day you review, instead of rereading the book, you pull the information out. Retrieval practice is a study method where you intentionally pull information out of your memory for the purpose of learning. This looks like closing the book and writing down what you can recall, or solving problems. This method was covered in "Retrieval Practice — The More You Pull It Out, the More It Stays". Workplace training for a little over 10,000 people that repeated sessions with time in between also used the method of solving problems and pulling information out (PMID 30623389).

When you pull information out, you mix the problems together. 155 high school students took a quiz every week for 4 weeks in a science class. The quizzes featured only half of the concepts learned that week (PMID 35436145).

On an exam one month later, students scored an average of 47% on concepts that were taught in class but did not appear on the quizzes. For concepts that appeared on quizzes where questions were grouped by single concepts, the score was 54%, and for concepts that appeared on quizzes where multiple concepts were mixed, it was 63% (PMID 35436145). Just pulling the information out was an improvement, and pulling it out mixed was an even bigger improvement.

When it comes to mixing, this article suggests a specific order. When background knowledge is low and difficult material is mixed right from the start, there was a result where blocking the material together appeared better on a test taken immediately after studying. After 8 weeks, that difference disappeared (DOI: 10.31234/osf.io/m46vf_v1). Therefore, this article suggests a careful sequence of first learning each type once before mixing them. There is no research in this article's evidence that directly tests this sequence itself.

And this calendar will feel difficult. The haziness on Tuesday and the hesitation in front of mixed problems mean that this method is working just as it is supposed to. It is like a cup of water given to a potted plant every day not thoroughly soaking the soil that very day.

Questions remain. Even an analysis showing that spaced repetition is effective in medical education notes that the optimal design and the long-term effects over longer periods require further study (DOI: 10.1111/tct.70353).

Given the same amount of time, space things out, pull information out when reviewing, and mix the problems up. This article suggests scheduling the mixing after you have learned each type separately. The measure of success is not how smooth the day feels, but what remains a month later.

Where to Look Next

This article covered when to study and in what order. To place spaced practice and interleaving side-by-side with other learning methods like practice testing, elaboration, and self-explanation, and to take a closer look at which methods have been verified to what extent, you can continue on to "The Science of Studying: What Actually Works".

Other articles in "Memory & Study from Scratch" lay the foundation for this one. How memories are put in, kept, and taken out is covered in "The Three Steps of Memory — Putting It In, Keeping It, Taking It Out", the difference between memories fading and simply not being able to be retrieved is covered in "Forgetting — What Fades and What We Just Can't Get Out", and calendar-based retrieval is covered in "Retrieval Practice — The More You Pull It Out, the More It Stays".

Source Papers

• "Distributed practice in verbal recall tasks: A review and quantitative synthesis." (2006) — DOI: 10.1037/0033-2909.132.3.354 — A meta-analysis of distributed practice across 184 papers and 317 experiments. The optimal learning interval for long-term retention grew longer as the time until the final test (retention interval) increased.
• "The Distributed Practice Effect on Classroom Learning: A Meta-Analytic Review of Applied Research." (2025) — PMID 40564553 — Across 22 classroom studies (31 effect sizes, over 3,000 participants), distributed practice outperformed massed practice by a moderate margin (d = 0.54). Longer retention intervals generally yielded larger effects (formal moderation analysis was precluded by a limited number of studies).
• "The spacing effect stands up to big data." (2019) — PMID 30623389 — In longitudinal data from 10,514 professional trainees, the interaction where the optimal interval between repeated retrievals increased alongside the retention interval matched laboratory findings.
• "Improving Students' Learning With Effective Learning Techniques: Promising Directions From Cognitive and Educational Psychology." (2013) — PMID 26173288 — An evaluation of 10 learning techniques. Practice testing and distributed practice showed high utility, while interleaved practice showed moderate utility as its evidence had just begun to be systematically explored.
• "Why Desirable Difficulties 'Work': A Review of the Evidence From Cognitive and Educational Psychology and Some Caveats for the Health Professions Education Field." (2026) — PMID 41508718 — Desirable difficulties hinder initial learning but aid long-term retention and transfer. Formative assessment, interleaving, and distributed practice have robust supporting evidence. The review cautions against indiscriminately equating greater difficulty with greater educational gain.
• "Desirable Difficulty: Theory and application of intentionally challenging learning." (2023) — PMID 35950522 — Difficulties requiring greater effort can impair short-term learning while benefiting long-term learning. The review explains retrieval, distribution, and interleaving through three theories.
• "Spaced retrieval: Absolute spacing enhances learning regardless of relative spacing." (2011) — DOI: 10.1037/a0023436 — Expanding the overall interval between repeated retrievals substantially increased long-term retention, but expanding, equal, and contracting intervals were inherently comparable in shape (laboratory experiments).
• "Interleaved practice improves mathematics learning." (2014) — DOI: 10.1037/edu0000001 — 126 seventh-graders practiced the same math problems using either interleaved or blocked practice for 3 months. Unannounced tests given 1 day and 30 days later both favored interleaving (d = 0.42, 0.79). Blocked homework reveals the solution method before reading the problem.
• "The effects of interleaved practice." (2009) — DOI: 10.1002/acp.1598 — Practicing four types of math problems using interleaved or blocked formats with fixed intervals. Interleaving reduced practice performance but doubled next-day test scores, and error analysis showed it enhanced the ability to match problems with the correct procedure.
• "A randomized controlled trial of interleaved mathematics practice." (2019) — DOI: 10.1037/edu0000367 — A preregistered cluster-randomized trial across 54 seventh-grade math classrooms over 4 months. On an unannounced test one month later, interleaving scored 61% versus blocked at 38% (d = 0.83). It concluded that while significant cues remain, the approach is effective and feasible.
• "Learning Painting Styles: Spacing is Advantageous when it Promotes Discriminative Contrast" (2011) — DOI: 10.1002/acp.1801 — Merely increasing the time interval between paintings by the same artist provided no benefit; interleaving or simultaneously presenting paintings by different artists improved the ability to identify the artist of unseen paintings. The advantage stems from contrast.
• "Similarity matters: A meta-analysis of interleaved learning and its moderators." (2019) — PMID 31556629 — A meta-analysis of 59 studies found an overall moderate interleaving effect (g = 0.42). The effect was largest for paintings, smaller for mathematics, unclear for expository texts and tastes, and favored blocking for vocabulary. The effect grew larger as categorical similarity increased. The authors concluded that while effective for inductive learning, materials and contexts should be considered, with caution advised for expository texts and vocabulary.
• "The limits of interleaving: How prior knowledge influences learning success in complex physics topics" (2025) — DOI: 10.31234/osf.io/m46vf_v1 — A field experiment with 176 high school students (preprint). Higher prior knowledge favored interleaving, whereas lower prior knowledge favored blocking, though this interaction vanished on an 8-week delayed test.
• "Interleaving in Mathematical Category Learning" (2025) — DOI: 10.5951/jresematheduc-2024-0055 — Pointing out that the rationale for recommendations confounded interleaving and spacing effects, two large classroom experiments learning angle relationships failed to replicate the interleaving versus blocking effect.
• "Scheduling math practice: Students' underappreciation of spacing and interleaving." (2022) — DOI: 10.1037/xap0000391 — Across two studies involving 193 and 175 college students respectively, the majority mistakenly judged schedules with the least spacing and interleaving to be the most effective, while viewing schedules with ample spacing and interleaving as more difficult, less enjoyable, and less common.
• "Perceiving effort as poor learning: The misinterpreted-effort hypothesis of how experienced effort and perceived learning relate to study strategy choice." (2019) — PMID 31470194 — Exposing participants to interleaving versus blocking (Study 1) and retrieval versus restudy (Studies 2 and 3) led them to rate strategies that felt harder as less effective and choose them less often. In Study 3, choosing the effortful strategy led to better long-term retention, contrary to their judgments.
• "Interleaving Retrieval Practice Promotes Science Learning." (2022) — PMID 35436145 — 155 students in grades 9–12 completed weekly quizzes over 4 weeks. On a test one month later, unquizzed concepts scored 47%, blocked-quiz concepts 54%, and interleaved-quiz concepts 63%.
• "The Effectiveness of Spaced Repetition in Medical Education: A Systematic Review and Meta‐Analysis" (2026) — DOI: 10.1111/tct.70353 — A meta-analysis of 13 studies involving 21,415 participants showed that spaced repetition outperformed standard study methods. Optimal design and longer-term effects require further research.