The 2-3-5-7 Study Method: Which Version Actually Works?

You saw the 2-3-5-7 study method on your feed, wrote the numbers on a sticky note, and then realised you had no idea what they counted. Days? Minutes? Days since what?

Fair question. The internet cannot agree either.

One popular article says study on day 1, then review on days 2, 3, 5 and 7. A university revision page says start from your exam date and count backwards. And at least one big study-tips page describes it as study blocks of 2, 3, 5 and 7 minutes. Same four numbers, three completely different plans.

I went looking for where the method came from and what the research says about each version. The idea behind it is one of the best-supported findings in learning science, the specific numbers are made up, and the most popular way of doing it quietly skips the part that makes spacing work.

What is the 2-3-5-7 study method?

The 2-3-5-7 study method is a spaced repetition schedule. You learn a topic once, then come back to it on a set pattern of later days instead of cramming it in one sitting. The numbers are the days you return.

If spaced repetition itself is new to you, this guide to spaced repetition covers the basics. Memories fade fast after you learn something, and each well-timed review slows that fade down. That's the forgetting curve at work, and 2-3-5-7 is one attempt to time reviews against it.

Where did the 2-3-5-7 method come from?

Nobody seems to know. I could not find a researcher, a paper or even a first post that introduced it.

Birmingham City University has a revision page that teaches the 2357 method and credits the underlying idea to Hermann Ebbinghaus, the 19th-century psychologist who first measured forgetting. The University of Kent has pushed it on TikTok. A widely shared 2025 article syndicated on Yahoo calls it "a modification of one of those older, time-tested techniques." None of them name who picked 2, 3, 5 and 7.

So the method is a real principle wearing a made-up number pattern. It looks like someone picked the first four primes because they're easy to remember. Honestly, that's fine. It just means the numbers themselves carry no scientific weight, and you should feel free to bend them.

The three versions of the 2-3-5-7 study method

Version 1: days counted forward

This is the one on most blogs. Day 1, learn the material. Then review on day 2, day 3, day 5 and day 7. Five sessions in a week, then you're "done."

Version 2: days counted back from the exam

The Birmingham City University version. Put a session on the day before your exam. Count two days back from that and add another. Then three days back, five days back, and seven days back for your first session. That puts your sessions 18, 11, 6, 3 and 1 day before the exam.

Version 3: 2, 3, 5 and 7 minute blocks

Adobe's study methods page describes the 2 3 5 7 method as "dividing study sessions into intervals of 2, 3, 5, and 7 minutes." Everything happens in one sitting.

Does the 2-3-5-7 study method actually work?

The spacing part does. Very well, in fact.

A 2021 meta-analysis in Educational Psychology Review by Hugo Latimier, Hugo Peyre and Franck Ramus pooled 29 studies on spaced retrieval practice, meaning testing yourself on the same material at intervals. Spacing those self-tests out beat doing them back to back by a large margin (an effect size of 0.74, which is big for education research).

But the same paper checked the other thing 2-3-5-7 assumes. Should the gaps get longer each time, the way 2, 3, 5, 7 does? The answer was no. Expanding gaps and equal gaps came out almost identical, an effect size of 0.034, which is basically zero.

That matches an older experiment. In 2007 Jeffrey Karpicke and Henry Roediger had students learn vocabulary pairs with either expanding or equal gaps between self-tests. Expanding gaps looked better on a test 10 minutes later. Two days later, the equal gaps won. Their conclusion was that what helps long-term memory is making the first recall come after a delay, so it takes real effort, not stretching each gap a bit more than the last.

So the growing sequence, the thing that makes 2-3-5-7 look clever, is the part with the least evidence behind it.

How each version holds up

Version 1 has a timing problem

Counting forward packs five sessions into your first week. If your exam is three weeks away, you then have a two-week hole with no reviews right before the moment that matters.

The best data on gap size comes from a 2008 study by Nicholas Cepeda and colleagues, published in Psychological Science. They taught more than 1,350 people a set of facts, gave them one review after a gap of up to 3.5 months, then tested them after a further delay of up to a year. The best gap depended on how far away the test was. For a test one week out, the ideal gap was roughly 20 to 40 percent of that week. For a test a year out, it dropped to 5 to 10 percent.

The right spacing depends on your exam date. A schedule that ignores the exam date is guessing.

Version 2 fixes the timing

Counting back from the exam anchors every session to the day you actually need the material. Your last session lands the day before, and the earlier ones stretch back far enough to give the memory time to settle. Of the three versions, this is the one I'd use.

Its weak spot is the word "revise." Neither the Birmingham page nor most blogs say what you do in each session, and for a lot of students "revise" means reread. Rereading on a spacing schedule is better than rereading once, but it wastes most of the benefit. The Latimier meta-analysis was about spaced retrieval, pulling the information out of your head, not looking at it again.

Version 3 is not spaced repetition

Two to seven minute blocks in one sitting is a focus or break trick at best. The Cepeda data says the helpful gap for a test a week away is measured in days, not minutes. Karpicke and Roediger's 10-minute advantage for expanding gaps faded by the two-day test. If your exam is next week, minute-scale spacing will make you feel prepared tonight and not much else.

The 2-3-5-7 study method, rebuilt

Keep the countdown from Version 2. Change what happens inside each session. Stop worrying about the exact numbers.

Pick one topic and your exam date. Say your biology midterm is on the 23rd. Count back 1, 2, 3, 5 and 7 days from the previous session each time. That gives sessions on the 22nd, 20th, 17th, 12th and 5th.
Session one is learning. Read the chapter or go through the lecture slides properly. This is the only session where reading is the main job.
Every later session starts with a closed book. Write everything you remember on a blank page, or answer practice questions cold. Only open your notes after you've tried. That's retrieval, and it's the step every popular version leaves out. Doing it across several days has its own name, successive relearning, and it's one of the strongest methods in the research.
Fill in the gaps, then stop. Check what you missed, fix it, and move on. A session can be 20 minutes.
Bend the schedule for weak spots. If you bombed the recall on the 12th, add an extra session a day or two later. A 2016 study by Everett Mettler, Christine Massey and Philip Kellman found that a schedule that adapted to each learner's accuracy and speed beat fixed schedules, expanding or equal. Your own misses are better data than four prime numbers.

Five topics for five exams? Stagger them so you're not running five countdowns that all hit the same afternoon. A retrospective revision timetable handles that well, because it tracks what you actually covered and how it went instead of a fixed plan.

If writing your own practice questions for every session sounds like a lot, a tool like StudyLab can turn your lecture PDFs into quizzes, which makes the closed-book step quick.

Common mistakes with the 2-3-5-7 method

Treating day 7 as the finish line. If the exam is weeks away, a forward-counted schedule ends too early. Count back from the exam.

Rereading in every session. It feels productive because the material looks familiar, but familiarity is not recall. Close the book first.

Following the numbers when you clearly forgot. The pattern is arbitrary. Add a session where you struggled.

Starting the countdown two days out. The backward version needs about 18 days. With less time, squeeze the gaps, but keep at least three spaced sessions with recall.

FAQ

Is the 2-3-5-7 method scientifically proven?

The spacing principle behind it is one of the best-supported results in learning research. The specific 2, 3, 5, 7 pattern is not. No study has tested those exact numbers, and a 2021 meta-analysis found that growing gaps work about as well as equal gaps.

What does 2-3-5-7 mean, days or minutes?

In the versions that match the spacing research, it means days. The minutes version you'll see on some pages is a focus technique and does not give memories the time between sessions that spacing needs.

How long should each 2-3-5-7 session be?

There is no fixed length. After the first learning session, 20 to 40 minutes of recall and fixing mistakes per topic is enough for most students. Stop when you've checked everything, not when a timer says so.

What if my exam is less than a week away?

Shrink the gaps and keep the recall. Three sessions spaced a day or two apart, each starting with a closed-book brain dump, will beat one long cram session. Just don't expect the same long-term retention you'd get from a longer runway.

Can I use the 2-3-5-7 method with flashcards?

Yes, and it pairs well. Flashcards force recall by design. Go through the whole deck on each session day, and pull the cards you missed into an extra mini-session before the next one.

Try it on one topic tonight

Pick the exam that's furthest away, write its date on a page, and count back 1, 2, 3, 5 and 7 days to get your five sessions. Put them in your calendar right now, before you lose the motivation. Tonight's session is the reading one. Every session after that starts with a blank page.

Want the recall step without writing the questions yourself? Upload the chapter to StudyLab and it'll generate a practice quiz for each session.