Yes, you can use spaced repetition on a book without making a single flashcard — because spacing and flashcards were never the same thing. Spaced repetition is the psychological effect; flashcard software is one implementation of it. Only two things actually have to be true:
- Retrieval. You have to try to pull the idea out of your head, before checking whether you got it right.
- Spacing. Those attempts have to be spread out over time, with gaps long enough that recalling takes effort.
Cards are one way to satisfy both. So are several others. Below: what the usual “without flashcards” advice really asks of you, a schedule you can run on a book today, and what changes when the prompts and the schedule are generated for you.
Can you do spaced repetition without flashcards?
You can, and conflating the two is the reason a lot of people give up on the idea entirely. As Hacking Chinese puts it, the two terms get used interchangeably but “the first is an expression of the spacing effect, the second usually refers to flashcard programs that make use of this effect” — so “if you don’t like Anki or Skritter, you should find some other way of making use of the spacing effect” (Hacking Chinese).
That distinction matters because Anki is genuinely excellent — free, open source, and the reason most people have heard of spaced repetition at all. The objection almost nobody voices precisely is not the card format. It is the authoring: turning a 300-page book into a deck is hours of work you do after the reading, on your own initiative, before any of the benefit arrives. Most people stall there, and conclude spaced repetition “didn’t work for them.”
The evidence, meanwhile, is about the mechanisms, not the tool. In their review of ten common study techniques, Dunlosky and colleagues rated practice testing and distributed practice as the two high-utility techniques — effective across ages, ability levels, and a wide range of material (Association for Psychological Science, Dunlosky, American Educator). Nothing in that finding requires index cards.
What “without flashcards” usually means online
Search this and you will get a list of alternatives. The most complete one recommends eight: two-column question-and-answer documents, hidden toggles in Notion, blurting (“reading a section of a textbook or your notes… and then writing down as much information as you can remember”), reciting aloud, the Feynman technique, problem sets, fill-in-the-blank diagrams, and mindmaps you redraw from memory — reviewed roughly 1 hour, 1 day, 2 days, 5 days, 10 days and 30 days after learning (Learning Rabbit Hole).
These are real techniques and they do work. Blurting in particular is close to pure retrieval practice, and it’s free.
But look at what each one asks of you. You choose the technique. You prepare the material. You decide the intervals. You keep the calendar. You remember to show up on day 10, a week after you last thought about the book. “Without flashcards” here means “with a different kind of manual work” — the labour moves, it doesn’t disappear. If the reason you bounced off Anki was the maintenance rather than the cards, none of these solve your problem.
There’s a second gap: none of this advice is about a specific book. It’s generic study technique. Nobody tells you what to do with the 400-page non-fiction book you finished last month and can already feel slipping.
A spaced repetition schedule for a book you’ve read
If you want to do it by hand, here is the version that respects both mechanisms, at the level of a book rather than a fact:
- Work chapter by chapter, not page by page. One chapter is roughly one idea-cluster — the right unit for a recall prompt. A book becomes 10–20 review items, not 400.
- Write the prompt, not the answer. When you finish a chapter, write down two or three questions it should have equipped you to answer — “why does the author think X fails?”, “what was the evidence for Y?”. This takes two minutes and creates no cards. The answers stay in the book.
- Use expanding intervals. Revisit at roughly day 1, day 3, day 7, then push out — day 21, day 60. The exact numbers matter far less than the shape: each gap longer than the last, long enough that recall is effortful.
- Answer from memory first, always. Say it aloud or write it out before you open the book. The effortful, slightly uncomfortable attempt is the part that does the work; checking is just feedback.
- On a miss, reset that chapter’s clock. Failed to recall a chapter on day 7? Reread that section and restart its intervals from day 1. Successes push the next review further out.
That’s the whole method, and it costs a few minutes per chapter plus the discipline to run a calendar for two months. Which is precisely where most people lose it — not because the method is wrong, but because nothing in the system reminds you, and nothing notices when you skip it.
Why highlighting isn’t the same as remembering
The most popular product answer to this problem is to highlight as you read and let a tool resurface the highlights later. Readwise does this well and deserves credit for it: it resurfaces the notes and highlights you take while reading, and its Mastery cards are “resurfaced according to a spaced repetition algorithm and presented in active recall form” — question-and-answer pairs and cloze deletion, with a next review typically one to four weeks out (Readwise). That is genuine retrieval on a genuine schedule, without you writing cards. If you already highlight heavily, it’s a good fit.
The structural limit is what it’s working from. A highlight-based system can only ever resurface what you highlighted — selected in the moment, while reading, before you knew which ideas would turn out to matter. The chapter you read attentively without marking anything is invisible to it. Your review coverage is a copy of your highlighting habits.
And highlighting is a weak instrument for this. The same review that rated practice testing and distributed practice as high utility rated highlighting and underlining, summarisation, and rereading among the lowest-utility of the ten techniques assessed (Association for Psychological Science). That isn’t an argument against tools built on highlights — the retrieval they perform is the high-utility part. It’s an argument against letting the selection step be the thing that decides what you’ll remember.
What it looks like when the schedule runs itself
The manual method above has two failure points: writing the prompts, and running the calendar. Both are automatable, and that’s what Study Junkie’s scheduled recall does.
You upload a book, PDF, DOCX or text file, and it’s turned into a course with auto-generated chapters from the full source — it preserves the material rather than summarizing it, so the recall prompts come from the actual book. Then:
- Recall quizzes are generated from your source material when the chapter content is created. Not from your highlights, and not hand-written by you — from the chapter itself, so coverage doesn’t depend on what you thought to mark.
- You choose the trigger: recall starts when you mark a chapter read, when you finish its assignments, or either.
- You set the schedule as day offsets — 1, 3, 7 works well, and it’s the same expanding shape as the manual version.
- You get reminded by email or push when a session is ready, and go straight into it. This is the part a calendar app doesn’t do: it arrives with the questions already made.
Around it sits the rest of the loop — audio narration of the chapters at 0.75×–2×, a tutor grounded strictly in your uploaded source, and assignments with AI feedback so you find out whether a chapter landed instead of assuming it did. We wrote about the feature itself in scheduled recall is here, and about how the chapters get built in turning any book or PDF into a course. For the wider system around retention, see how to actually remember what you read.
Honest caveats: this costs money where blurting is free, and it can’t make you show up — it can only make showing up take two minutes instead of twenty. If you’re working through one book with high motivation and no budget, the manual schedule above is a genuinely good answer. If you have a stack of them and know from experience that the calendar is where this dies, automating the prompts and the timing is the difference between a system you have and a system you use.
See how the whole loop works, or start with 1,000 free Study Credits — no subscription, and credits don’t expire, so they wait until you have a book worth studying.
FAQ
Can you do spaced repetition without flashcards?
Yes. Spaced repetition is the spacing effect; flashcards are one implementation of it (Hacking Chinese). Any method that makes you retrieve information from memory at increasing intervals qualifies — spoken recall, blurting, chapter questions, problem sets, or auto-generated recall quizzes.
What is the best spaced repetition schedule for a book?
Expanding intervals at the chapter level: roughly day 1, day 3 and day 7, then stretching to day 21 and day 60. Reset a chapter to day 1 whenever you fail to recall it. The precise numbers matter much less than the pattern of each gap being longer than the last.
Is highlighting a book enough to remember it?
No. Highlighting and underlining were rated among the lowest-utility techniques in Dunlosky and colleagues’ review of ten study methods, while practice testing and distributed practice were the two rated high utility (Association for Psychological Science). Highlights are useful as a record of what caught your attention; remembering requires retrieving.
How is this different from just rereading the chapter?
Rereading feels productive because the text is familiar, but familiarity isn’t recall. Rereading was also rated low utility in the same review, while retrieval practice was rated high. Trying to answer before you look is the part that builds the memory.
Do I have to make Anki cards to use spaced repetition?
No, and Anki being excellent doesn’t change that. The barrier for most readers is the authoring time — turning a long book into a deck before any benefit arrives. Chapter-level prompts you write in two minutes, or recall quizzes generated from the source automatically, satisfy the same two requirements with far less setup.