Spaced repetition schedule generator

How to use the generator

1. Type the topic into “What you’re learning”. It becomes the title of each calendar event, such as “Review: Biology chapter 3 (2 of 4)”.
2. Set “First study date”. It’s today unless you change it.
3. Under “Plan for”, choose “A test date” and set “Test date”, or choose “Long term” and pick a period under “For how long”.
4. Set “Number of reviews”, from 1 to 8.
5. Select “Download calendar file (.ics)” to import the reviews as all-day events into a calendar app, or “Copy as text” to paste the list into notes or a message.

Each calendar event says the same thing: close your notes and test yourself first. Nothing you type leaves your browser; the file is built on your device.

How the dates are worked out

The generator follows three steps.

1. Find the ratio. Cepeda and colleagues found that the best gap before a single review was about 20 to 40% of the time from that review to the test when the test was a week away, falling to about 5 to 10% when it was a year away. We use the midpoints, 30% and 7.5%, and blend between them on a log-time scale. Outside the week-to-year span the nearer value holds.
2. Set the first gap. The first review is placed so that its gap is that share of the time left until the test. As a formula: first gap = r × T ÷ (1 + r), where T is the days from first study to the test and r is the ratio.
3. Grow the gaps. Each later gap is longer than the one before, by the same factor, and the longest gap runs into the test. If your reviews only fit with gaps growing by less than 1.5 times, the first gap is shortened instead, never below 1 day.

Here’s the default. You start on Saturday, October 17, 2026, the test is 30 days later on Monday, November 16, and you ask for 4 reviews. For 30 days the ratio is about 18%, so a single review would land on day 5 (5 of the remaining 25 days is 20%). Four reviews don’t fit behind a 5-day first gap, so the gap shortens to 2 days and the reviews fall on days 2, 6, 11 and 18: Monday, October 19; Friday, October 23; Wednesday, October 28; and Wednesday, November 4. The gaps are 2, 4, 5 and 7 days, then 12 days to the test.

The result card also shows a model check. It runs the same dates through FSRS-6, an open memory model used by flashcard apps, with its published default settings. For the default schedule it gives about 98% recall on test day, against about 67% with no reviews (our calculation). That number assumes you answer every review correctly, and it describes an average flashcard, not you.

How long to wait between reviews

A 2006 review by Cepeda and colleagues pulled together 839 assessments of distributed practice from 317 experiments in 184 articles. The best gap between study sessions increased as the time to the final test increased. The authors noted that few studies had tested memory a day or more later.

Their 2008 study filled that gap with more than 1,350 people. Participants learned trivia-style facts, reviewed them after a gap of up to 3.5 months and took a final test up to a year later. The best gap was about 20 to 40% of a 1-week test delay and about 5 to 10% of a 1-year delay. Put simply, the further away the test, the longer you can wait before the first review, though the gap shrinks as a share of the total.

Test yourself at each review

The way you review is a separate question from when. In Roediger and Karpicke’s 2006 experiments, students read short prose passages. In the second experiment, those who studied once and then took three recall tests remembered 61% a week later; those who studied three times and took one test, 56%; those who studied four times, 40%. The four-study group was the most confident it would remember. So at each review, close the book and write or say what you can recall, then check.

Cramming and spacing

A 2013 review of 10 study techniques rated only two as high utility: practice testing and spacing out study. Rereading and highlighting were rated low utility, which doesn’t make them useless. In a math experiment with 66 college students, splitting the same practice across two sessions a week apart lifted scores on a test a week later from 49% to 74%, even though both groups scored about the same during practice.

Fixed ladders and long-term plans

Some study tools use a fixed ladder, such as reviews 1, 3, 7, 14 and 30 days after learning. It has the same shape as this generator’s long-term mode (gaps that grow) and is fine when there’s no test date. The generator’s advantage is that it scales the gaps to your date: a test 9 days away and a test 90 days away get different plans. If a flashcard app already schedules your cards, let it; this tool is for material that lives outside one, such as a textbook chapter, lecture notes or a talk. For a picture of what each review does to modelled recall, try the forgetting curve calculator.

What the schedule doesn’t tell you

The dates are a starting point. Cepeda’s ratios come from trivia-style facts and a single review, so a dense subject, a skill or a language may need more reviews or different gaps. The schedule can’t see how well you learned the material on the first day or how each review went. The model check is an average for flashcards that assumes you get every review right, so read it as a comparison between plans, not a forecast of your score.