Leitner System vs Anki for Exam Prep

If your deck will stay at a few hundred cards and you remember things better when you write them by hand, a paper Leitner box is enough. If you face a syllabus of thousands of cards, such as the MCAT or the bar exam, use Anki. The difference is how reviews get timed. Leitner gives every card in a box the same review day. Anki gives each card its own date.

Both methods use spaced repetition, which means reviewing material at longer and longer intervals so it moves into long-term memory. Both also use retrieval practice: you look at the front of a card, try to recall the back, and only then check. The open question is who decides when a card comes back. With Leitner, a box decides. With Anki, a scheduling algorithm decides, and that algorithm is either SM-2 or FSRS, explained below.

Sebastian Leitner introduced the method in 1972. You write flashcards and sort them into boxes. The University of York's study guide recommends 3 to 5 boxes. A card you answer correctly moves up one box. A card you miss goes back to Box 1, wherever it was. Box 1 gets reviewed every day, Box 2 every other day and Box 3 once a week. Other descriptions use intervals of 1, 2, 4 and 8 days. Our page on the schedule for 3, 5 and 7 boxes lays out each version day by day.

The method is good at sending your time to the cards you keep missing, because those are the cards stuck in Boxes 1 and 2. The boxes also show your progress. You can see the pile moving right. Research summarized by Tabibian and colleagues in 2019 described the method as having "achieved reasonable success in practice." Whether it matches a modern algorithm card for card is a separate question, and our evidence check on Leitner covers it.

The weaknesses come from the same simplicity. Every card in a box is treated the same, so a card you barely remembered and a card you knew instantly come back on the same day. Grading is pass or fail, so there is no way to say "I got it, but it was hard." The York guide warns that managing many cards across several subjects can become overwhelming, and that making and sorting all the cards takes a lot of time up front.

It costs a few dollars. Buy it if your deck is small, you learn by writing, or you need to study without a screen.

Anki is free software that has scheduled flashcards since 2006. It does not use boxes. Every card carries its own review date, and you grade each answer with one of four buttons. The date comes from one of two schedulers.

SM-2 is a formula Piotr Wozniak published in 1987. It gives each card an ease factor, a multiplier that sets how fast its interval grows, starting at 2.5. Anki's older version of it has a floor of 1.3. A card you keep missing can get stuck near that floor and come back again and again without ever graduating.

FSRS, the Free Spaced Repetition Scheduler, started as an open-source project in 2022. It estimates how likely you are to remember each card and schedules the review for the day that likelihood drops to a target you choose. One explainer recommends 90% as a sensible target. The Anki manual warns that targets above 97% "can be overwhelming," because the number of reviews rises steeply as you aim for near-perfect recall.

As of September 2026, the current version is Anki 26.09, released September 14, 2026. It updated FSRS to 6.6.2, up from the 6.1.x line that the FSRS project's release notes say has been supported since Anki 25.09. FSRS is still a setting you turn on or off: the 26.09 changelog includes fixes that apply "when FSRS is disabled." The sources disagree on which scheduler is the default. One says FSRS became the default in version 23.10, and another says SM-2 is still the default. The reliable check is to open your deck's options and look. Version 26.09 also patched two security holes, and a 26.09.2 patch arrived the next day. If you install Anki now, get 26.09 or later. Our page on the latest Anki security fix covers the upgrade.

Anki's biggest advantage for exam prep is what other people have already built. There are 100,000+ shared community decks, and one survey-based figure puts Anki use among American medical students above 86%.

It is free on desktop and Android and costs $25 one time on iOS. Buy it if your exam requires thousands of facts or if a good shared deck already exists for it.

Three things separate these methods. Everything else is a matter of taste.

First, the number of cards. SmartRecall puts Leitner's comfortable limit at 200 to 500 cards. Above that, review sessions become hard to manage, because a single box can hold hundreds of cards that all come due on the same day. Anki stays manageable past 10,000 cards because cards come due one at a time. That is roughly the gap between learning one chapter of vocabulary and covering an MCAT or bar syllabus.

Second, how precise the timing is. A Leitner box is like a class that moves at one pace. Anki is like a tutor who sets a pace for each card. The claim that this saves time comes mostly from vendors. StudyGlen reports that FSRS needs 20 to 30% fewer reviews than SM-2 for the same retention, based on benchmarks from 500+ million Anki reviews. StudyGlen sells a product built on FSRS, and none of these sources compare FSRS directly with paper Leitner boxes. A head-to-head test, with the same learners, the same cards and the same exam, would settle the question. That study is not among the sources here.

Third, how much data the scheduler needs. FSRS personalizes its schedule from your own review history, and estimates say this takes about 400 to 1,000 reviews before it beats the default settings. Below that, FSRS performs about the same as SM-2. For a small deck studied for a few weeks, the algorithm's main advantage never fully kicks in.

One point cuts across all three. Mindomax's reading of the research is that "consistency matters more than algorithm choice." Meta-analyses it cites found large benefits from spaced retrieval (Latimier and colleagues, 2021) and from practice testing (Adesope and colleagues, 2017). Both methods give you those benefits if you show up every day. A paper box you open daily will beat an Anki deck you avoid.

Use a paper Leitner box if your deck is under a few hundred cards, you remember better by writing, or you want study time away from a phone. A vocabulary list for one class, the formulas for a single unit, and drug classes for one block are good fits.

Use Anki if your exam involves thousands of cards, if a well-kept shared deck already exists for it, or if you want each card timed on its own rather than by the box. For the MCAT or the bar, that will usually be the right choice from the start, so you avoid copying everything over later.

A rule of thumb for switching: move from paper to Anki once your deck passes about 500 cards, or once Box 1 alone takes longer than you are willing to spend each day. When you switch, use Anki 26.09 or later, and check in deck options which scheduler you are on. Expect FSRS to work much like SM-2 until you have logged several hundred reviews.

If you are staying with paper, the next step is building the box itself.