Spaced repetition in Medicine: practical guide with Anki

In a study with more than 1,350 people, the best interval between study and review grew as the test got further away (Cepeda et al., 2008). But, as a fraction of the time until the test, it shrank: from about 20 to 40% for a test in 1 week to about 5 to 10% for a test in 1 year (Cepeda et al., 2008). In other words, the timing of the review matters, and what is good for next week's test is different from what is good for remembering during the internship.

NEXO is a daily medical diagnosis game: one clinical case a day, six progressive clues, and ICD-10 proximity feedback across five levels.

This text is a practical guide to spaced repetition for those who study Medicine with Anki: what spacing is, what studies in medical education show and what they still do not show, how to write good cards, how to control the volume of reviews and how to take NEXO cases to Anki. The general principle of reviewing at intervals appears in the text about how to fit games into your routine, and the testing effect, which is the other half of spaced repetition, is in the text about reviewing by playing.

Key points

• Spacing out review helps retention, and the ideal interval depends on when you need to remember.
• In medical education, spaced questions and cases improved retention in randomized trials.
• Studies on Anki are observational: there are associations with higher grades and also a recent study with no association.
• Good cards ask you to produce the answer, test one idea at a time and come from your mistakes.
• The volume of reviews grows with new cards; adjust new cards to what you can review.
• In NEXO, the case is the test, and the Anki button takes what you got wrong to your spaced review.

Summary

• Spacing
• Evidence in medical education
• How Anki schedules reviews
• Good cards
• Volume
• The Anki button in NEXO
• A routine with Anki
• Frequently asked questions
• Spaced repetition checklist

Spacing

Spaced repetition is returning to the same content in sessions separated in time, instead of concentrating all repetitions at once. In Anki and similar tools, each return is a small test: you see the question, try to remember and only then check.

A term with many meanings

In health education, "spaced learning" means quite different things. A scoping review included 120 articles on the topic and concluded that the term is poorly and vaguely defined in the literature of the field (Versteeg et al., 2020). The formats ranged from distributing content over a single day to intervals of several months (Versteeg et al., 2020). In most cases, spacing was applied in the practice or testing phases, and rarely during teaching (Versteeg et al., 2020).

So when a study says that "spacing worked", it is worth asking: spaced what, with what interval and until when? The answer changes what you can take into your routine.

How long to wait between one review and another

The classic meta-analysis on distributed practice, already summarized in the text about study routine, showed that the interval between sessions and the time until the test act together on retention. The 2008 study measured this on a larger scale: people learned a set of facts, did a review after an interval of up to 3.5 months and a final test after up to 1 year (Cepeda et al., 2008).

For any test deadline, increasing the interval until the review first improved and then gradually worsened the final result (Cepeda et al., 2008). That is, there is a good point: reviewing too early yields little, and waiting too long also does. The authors conclude that the interaction between interval and test deadline makes many educational practices very inefficient (Cepeda et al., 2008).

This is a study with simple facts, outside Medicine. It serves to understand the principle, and not to derive an exact review calendar from it.

Test of the week or knowledge for the internship

The same study helps to think about two different goals. For an upcoming test, the optimal interval was a larger fraction of the deadline; to remember after 1 year, the optimal interval was larger in days, but a smaller fraction of the deadline (Cepeda et al., 2008). In Medicine, many things studied in the basic cycle come back in the clinical cycle, in the internship and in the residency exam.

The Anki scheduler handles this automatically: cards you get right come back at increasingly longer intervals, and the ones you get wrong come back soon (Anki manual). What is up to you is keeping the important cards in the deck after the exam, instead of deleting them as soon as it is over.

Spacing and testing go together

A review applied to medical education concludes that tests should be frequent and spaced over time to improve retention (Larsen et al., 2008). A review written for medical students brings together three methods that add up: the testing effect, active recall and spaced repetition (Augustin, 2014). In practice, an Anki card combines the three things: it is a test, it asks you to retrieve the answer from memory and it comes back at intervals.

Evidence in medical education

The best data on spaced repetition in Medicine come from trials with questions and cases sent at intervals. Studies on Anki itself are another matter: they measure the use students already make of it and look for association with grades.

Spaced questions and cases

In a randomized trial with third-year students, each one received, after the urology clerkship, weekly cases and questions by email about only 2 of the 4 topics of the clerkship (Kerfoot et al., 2007). Of the 133 students, 95 took the end-of-year exam (Kerfoot et al., 2007). The effect was greater in those who had done the clerkship longer ago: effect size of 1.01 for 6 to 8 months and 0.73 for 9 to 11 months (Kerfoot et al., 2007).

These were multiple-choice questions by email, with answer and explanation, and not Anki cards (Kerfoot et al., 2007). The authors conclude that cases and questions distributed every week can greatly improve retention of medical knowledge (Kerfoot et al., 2007). The general idea of this trial also appears in the text about study routine.

With residents, and with transfer

With urology residents, a trial compared a daily question by email, with a histopathology image and immediate feedback, with block modules of the same content (Kerfoot et al., 2010). Each spaced education cycle lasted 4 weeks and had 20 questions with different images, repeated 3 times (Kerfoot et al., 2010). The results on retention and adherence are in the text about how to study during residency.

Two points from the same trial are of interest to those who make cards. The authors report that spaced education multiplied the efficiency of long-term learning by 4 (Kerfoot et al., 2010). And they conclude that it generated transfer of the histopathological diagnosis skill, besides improving retention (Kerfoot et al., 2010).

What studies on Anki measure

In a study with 72 students from a school in the United States, all reported doing exam-style questions, with an average of 3,870 questions, and 31% used Anki (Deng et al., 2015). The number of unique Anki cards seen and the number of questions done were independent predictors of the licensing exam score (Deng et al., 2015). The number of cards seen in a commercial bank of ready-made cards, however, did not predict the score (Deng et al., 2015). The size of this association is in the text about types of games for medical students.

In a questionnaire at another school, Anki use was associated with higher scores on the same exam, and, among users, more consistent use was associated with higher scores and a greater sense of retention (Lu et al., 2021). The authors conclude that Anki is effective, but the design, a self-reported questionnaire, shows association, not cause.

A recent study with no association

In a cross-sectional study with 176 first-year students from a school in the United States, 82% used Anki and 51% had it as their main method (Spence et al., 2026). There, Anki use was not associated with better performance on internal exams (Spence et al., 2026). The authors conclude that success early in the course seems more linked to well-being and focused study habits than to a specific tool (Spence et al., 2026). The full text is under embargo; the information here comes from the abstract.

A recent narrative review on Anki in medical education points out the main limitation of this literature: small samples and self-reported data (Mohamed et al., 2025). The authors call for larger, prospective studies on long-term retention and performance (Mohamed et al., 2025).

How to read these results

Study	Design	What it found
Kerfoot et al., 2007	Randomized trial, third-year students	Spaced questions improved retention at the end of the year (Kerfoot et al., 2007)
Kerfoot et al., 2010	Randomized trial, residents	Daily spaced question outperformed block modules (Kerfoot et al., 2010)
Deng et al., 2015	Observational, one school	Unique Anki cards and questions associated with grade (Deng et al., 2015)
Lu et al., 2021	Questionnaire, one school	Anki use and consistency associated with grade (Lu et al., 2021)
Spence et al., 2026	Cross-sectional, one school, internal exams	Anki use with no association with performance (Spence et al., 2026)

The honest reading is this: the principle of testing at intervals has experimental support, including in medical education. Anki is one way to apply that principle, and studies on it show associations, with small samples and self-reported data (Mohamed et al., 2025). If you research this kind of question, NEXO Researcher is the program for proposing studies with NEXO.

How Anki schedules reviews

Anki is a card program with spaced repetition, maintained by third parties. This section describes how it works according to the official manual, so you understand what happens when you press each button.

The four buttons

After seeing the answer, you choose between Again, Hard, Good and Easy (Anki manual). The manual calls for rigor: if a partially correct answer would count as wrong outside Anki, it is also wrong in Anki, and Again applies (Anki manual). Hard is for a correct answer with doubt or delay; Good, for a correct answer with some effort; Easy, for a correct answer with no effort (Anki manual).

The manual describes as typical using Again about 5 to 20% of the time and Good about 80 to 95% (Anki manual). Anyone who finds four buttons hard can use only Again and Good (Anki manual).

Ten seconds to remember

According to the manual, it is fine to take a while to remember, but, as a general rule, if you cannot answer in about 10 seconds, it is better to show the answer and move on (Anki manual). This connects with what is known about testing: feedback after trying to remember is crucial for learning (Larsen et al., 2008).

Desired retention

In Anki's FSRS scheduler, desired retention is the proportion of cards you should remember when they come back; the default is 90% (Anki manual). Higher retention leads to shorter intervals and more reviews per day, and the manual warns that, above 90%, the load rises very fast (Anki manual).

The manual also warns about a habit that disrupts scheduling: pressing Hard when you forgot (Anki manual). Hard counts as correct, and, if used when you forgot, it leaves intervals too long (Anki manual). When you forget, use Again.

When you fall behind

If reviews pile up, Anki prioritizes, by default, the cards that have been waiting longest, and takes the delay into account when calculating the next interval (Anki manual). So, after a break, you pick up where you left off, without having to start from scratch.

Good cards

A good card is a good test. The suggestions below follow the principles of the studies and the manual, and those that are an editorial choice appear as a practical suggestion.

Ask, do not show

Tests that require effortful recall, such as short-answer ones, yield more retention than recognition ones (Larsen et al., 2008). For the card, this means a front that asks a question, not a front that already carries the answer hidden in the text. The test formats that yield the most are in the text about reviewing by playing.

One idea per card

As a practical suggestion, each card should ask for only one thing. A card that asks for "everything about the disease" mixes several answers, and you end up pressing Good because you remembered part of them. The manual suggests making cards simpler if you systematically go over 60 seconds per card (Anki manual).

Instead of	Try
Front: the name of the disease; back: a whole paragraph	Front: which exam confirms the disease?; back: the exam
Front: a text with the answer highlighted	Front: a question; back: the short answer
Front: "causes of the disease?"; back: a long list	One card per cause, each with the clue that leads to it
Card copied from a summary you did not study	Card written after missing a case or a question

Cards that come from your mistakes

In the study with 72 students, what was associated with the grade was the number of unique Anki cards seen; the number of cards seen in a commercial bank of ready-made cards was not (Deng et al., 2015). The study does not explain why, and you cannot claim that writing your own cards was the cause. As a practical suggestion, the most useful starting point for a card is what you just missed: the mistake shows the gap, and the card takes that gap into spaced repetition.

The feeling of knowing deceives

In a trial with electronic Nephrology cards, students' predictions about their own performance did not correlate with actual performance (Schmidmaier et al., 2011). In the same trial, after 6 months overall recall was low, and retesting and restudying tied (Schmidmaier et al., 2011). The short-term result is in the text about types of games. For your routine, the lesson is twofold: trust the test more than the impression that you already know it, and return to the card over the months.

Cards you always miss

Anki calls a card you repeatedly forget a leech; by default, when a card fails 8 times in review, Anki flags the note and suspends the card (Anki manual). The manual suggests three ways out: rewrite the card, because maybe it has too much information or you are trying to memorize something you did not understand; delete it, if the content is not worth the time; or wait, when the confusion comes from two similar contents learned together (Anki manual).

In Medicine, that last situation is common: two diseases with similar names or presentations, two drugs of the same class. As a practical suggestion, when two cards get confused, go back to the book or the case to understand what separates the two, and create a card that asks exactly that difference. Reviewing similar examples mixed together, instead of in blocks, is the topic of the section on mixing similar topics in the routine text.

Image on the card

For visual content, such as slides, imaging exams and clinical signs, as a practical suggestion, put the image on the front and the question about it, and leave the answer on the back. That way the card trains recognizing the finding, and not just its name.

Volume

There is no number of cards per day with good evidence for medical students. What exists are two useful things: how scheduling works, which explains where the load comes from, and some studies on hours and study methods.

Where the review load comes from

New cards generate reviews in the following days. The manual gives an example: someone who learns 20 new cards per day, steadily, can expect something like 200 reviews per day (Anki manual). The same manual reports users who studied hundreds of new cards in the first days and then became overwhelmed with the reviews (Anki manual). The example serves to understand the proportion, and not as a goal.

Anki has limits for new cards per day and reviews per day (Anki manual). When there are accumulated overdue reviews, the manual recommends stopping the introduction of new cards until you catch up on the queue (Anki manual). The Anki simulator lets you estimate the load in reviews or minutes per day for different settings (Anki manual).

More hours and more methods

In the cross-sectional study with first-year students, study hours did not correlate with results, and those who used three or more study methods had lower grades on written exams than those who used two or fewer (Spence et al., 2026). In the same study, more self-reported well-being and at least 7 hours of sleep were associated with better performance on written exams (Spence et al., 2026). These are associations from a study at one school, and not rules, but they remind us that more volume does not always mean more learning.

How to adjust your volume

As a practical suggestion:

• Start with what you can review, not with what you would like to learn. Tomorrow's load depends on today's new cards.
• Prioritize reviews over new cards when the queue grows.
• Cut cards that no longer serve you, such as those from a subject whose exam has already passed and that you will not need again.
• Use the simulator in Anki before changing the desired retention or the new card limit.
• Watch your sleep and your tiredness: if Anki is taking the place of rest, it is worth reducing new cards.

The Anki button in NEXO

In NEXO, each case is a short test: up to six clues, one at a time, and a guess with a free response, searching for the diagnosis in ICD-10. With each guess, the proximity appears in five levels, and the explanation at the end of the case shows what was missing. It is at that moment, right after the error, that the card gains meaning.

The Anki button creates the card straight from NEXO and opens it in the Anki app, searches for the cards you already have and, when the case has an image, takes the image to the card. Thus, today's case enters your spaced repetition, and Anki takes care of the intervals.

Daily, Archive and regularity

The Daily brings one case per day, the same for everyone, and renews at midnight: a new test every day, alongside Anki reviews. The Archive gathers previous daily cases and helps you review an area before the exam. The streak records the days you played, and the achievements mark the path. To understand the limits of each mode, see the NEXO economy page.

From case to card, step by step

As a practical suggestion:

1. Play the case and note the clue at which you changed your mind.
2. Read the explanation at the end of the case.
3. Create the card with the Anki button, with a question about the point that was missing.
4. Let Anki schedule the reviews and answer rigorously.
5. Return to cases from the same area in the Archive days later.

To research, bring to your class and create

If you research learning and want to propose studies on review and memory with NEXO, check out NEXO Researcher. To bring NEXO to your medical school, student league or residency, there is NEXO Closer. Doctors who review cases in their specialty take part through NEXO Curator, and those who create health content on social media can take part through NEXO Creators.

A routine with Anki

The routine below is a practical suggestion from the Editorial team, to adjust to your stage of the course and your time.

Moment	What to do
Every day, first	Anki reviews that are due
Every day, afterwards	Daily case; create cards only for what you got wrong
After class	New cards about what you studied and tested
Once a week	Cases from one area in the Archive and questions in the exam format
Before the exam	More tests on the topic and fewer new cards
After the exam	Review what you got wrong and cut the cards that no longer serve you

Common mistakes

• Creating too many cards at the start: the review load for the following days grows along with it (Anki manual).
• Pressing Hard when you forgot: the intervals become too long (Anki manual).
• Cards that show instead of asking: recalling with effort pays off more than recognizing (Larsen et al., 2008).
• Trusting the feeling of knowing: it may not track real performance (Schmidmaier et al., 2011).
• Reading an association study as proof of effect: studies on Anki are observational (Mohamed et al., 2025).
• Insisting on a card that never sticks: rewrite it, delete it or postpone it, as the manual suggests (Anki manual).
• Staying only with cards: questions and cases test the same content from another angle.

Frequently asked questions

Does Anki work for Medicine?

The principle it applies, testing at intervals, has support in trials in medical education (Kerfoot et al., 2007). Studies on Anki itself are observational: some associated its use with higher grades (Lu et al., 2021), and a recent study found no association with internal exams (Spence et al., 2026).

How many new cards per day?

The ideal number still lacks good evidence. Adjust new cards to the number of reviews you can do, remembering that each new card generates reviews in the following days (Anki manual).

How long should you wait between reviews?

It depends on when you need to remember: the optimal interval grows with the time until the exam (Cepeda et al., 2008). In Anki, the scheduler calculates the intervals from your answers.

Cards or questions?

Both. In the study with 72 students, the number of questions done and the number of unique Anki cards seen were independent predictors of the grade, each controlling for the other and for other academic factors (Deng et al., 2015). These are associations, not an effect measured in a trial. As a practical suggestion, use questions to apply the content in a prompt and cards for the facts and concepts you need to have at hand.

Is it worth using Anki during the clinical clerkship?

With little time, short and spaced units also worked for residents (Kerfoot et al., 2010). How to fit review into the breaks of the rotation is in the text about how to study during the clinical clerkship. As a practical suggestion, during the clerkship, create cards from the patients and cases of the day, and reduce new cards on topics that will not come up.

Ready-made deck or your own cards?

The studies did not test this directly. In the study with 72 students, the unique Anki cards were associated with the grade, and those from a commercial bank of ready-made cards were not (Deng et al., 2015). As a practical suggestion, if you use a ready-made deck, add cards from your mistakes.

I fell behind on reviews. Do I start over?

You can continue from where you stopped. Anki prioritizes the cards that have been waiting the longest and takes the delay into account in the next interval (Anki manual). Stop new cards until you get the queue up to date (Anki manual).

Is spaced repetition only for facts?

The spaced education trials used clinical cases and questions, not only facts (Kerfoot et al., 2007). To train reasoning with whole cases, combine cards with cases and questions. The spaced cases with images for residents also went beyond retention, according to the authors, with transfer of diagnostic skill (Kerfoot et al., 2010).

How does NEXO fit into spaced repetition?

The NEXO case is the test, and the Anki button takes what you got wrong to your cards, with the case image when there is one. The Daily mode keeps one new case per day, and the Archive lets you return to an area.

What is FSRS?

It is an Anki scheduler that estimates how much you tend to forget and sets the reviews from a desired retention, with a default of 90% (Anki manual).

Spaced repetition checklist

1. Do your cards ask questions, rather than show the answer?
2. Does each card cover one idea only?
3. Are the cards born from your mistakes in cases and questions?
4. Do you answer with rigor, using Again when you forget?
5. Does the review load fit into your day?
6. Do you stop new cards when the queue falls behind?
7. Do you test the same content also in cases and questions?
8. Do you return to the topics over the months, not only before the exam?
9. Do you read the studies on Anki as associations?
10. Do you take care of sleep even in exam week?

References

1. Cepeda NJ, Vul E, Rohrer D, Wixted JT, Pashler H. Spacing effects in learning: a temporal ridgeline of optimal retention. Psychol Sci. 2008. https://pubmed.ncbi.nlm.nih.gov/19076480/
2. Versteeg M, Hendriks RA, Thomas A, Ommering BWC, Steendijk P. Conceptualising spaced learning in health professions education: A scoping review. Med Educ. 2020. https://pubmed.ncbi.nlm.nih.gov/31860936/
3. Larsen DP, Butler AC, Roediger HL 3rd. Test-enhanced learning in medical education. Med Educ. 2008. https://pubmed.ncbi.nlm.nih.gov/18823514/
4. Augustin M. How to learn effectively in medical school: test yourself, learn actively, and repeat in intervals. Yale J Biol Med. 2014. https://pubmed.ncbi.nlm.nih.gov/24910566/
5. Kerfoot BP, DeWolf WC, Masser BA, Church PA, Federman DD. Spaced education improves the retention of clinical knowledge by medical students: a randomised controlled trial. Med Educ. 2007. https://pubmed.ncbi.nlm.nih.gov/17209889/
6. Kerfoot BP, Fu Y, Baker H, et al. Online spaced education generates transfer and improves long-term retention of diagnostic skills: a randomized controlled trial. J Am Coll Surg. 2010. https://pubmed.ncbi.nlm.nih.gov/20800189/
7. Deng F, Gluckstein JA, Larsen DP. Student-directed retrieval practice is a predictor of medical licensing examination performance. Perspect Med Educ. 2015. https://pubmed.ncbi.nlm.nih.gov/26498443/
8. Lu M, Farhat JH, Beck Dallaghan GL. Enhanced Learning and Retention of Medical Knowledge Using the Mobile Flash card Application Anki. Med Sci Educ. 2021. https://pubmed.ncbi.nlm.nih.gov/34956708/
9. Spence LE, Khoury SK, Gronewold A, Regan MP, Daly DT, Nebel C. Wellness and Work: Anki, Study Habits, and Exam Outcomes in First-Year Medical Students. Med Sci Educ. 2026. https://pubmed.ncbi.nlm.nih.gov/41939075/
10. Mohamed NSA, Alrafi MA, AlBani B, et al. Application of Anki Flashcards in Medical Education: A Narrative Review. Med Sci Educ. 2025. https://pubmed.ncbi.nlm.nih.gov/41798361/
11. Schmidmaier R, Ebersbach R, Schiller M, et al. Using electronic flashcards to promote learning in medical students: retesting versus restudying. Med Educ. 2011. https://pubmed.ncbi.nlm.nih.gov/21988625/
12. Anki Manual. Leeches. https://docs.ankiweb.net/leeches.html
13. Anki Manual. Studying. https://docs.ankiweb.net/studying.html
14. Anki Manual. Deck Options. https://docs.ankiweb.net/deck-options.html