How to Remember What You Read: Research-Backed Techniques

Reading Guides
By BookBits Editorial
How to Remember What You Read: Research-Backed Techniques

Forgetting most of a book is normal, and rereading it feels more useful than it is. If you want to know how to remember what you read, three habits have the strongest evidence behind them: test yourself instead of rereading, space your reviews over days and weeks, and explain ideas in your own words. Below you will find what each study actually tested, honest answers to the questions people ask (including the 2-7-30 rule), and a simple routine for your next book.

One caveat before we start. Most of this research used students, short passages, word lists and lectures, not whole books. For nonfiction readers it is the best evidence available, not a tested recipe, and we point out the limits as we go.

Why you forget most of what you read

The forgetting curve is real, and it has been measured again more than a century later. In a 2015 study in PLoS ONE, Murre and Dros replicated Ebbinghaus’s original 1885 experiment. One person, aged 22, spent 70 hours learning lists of nonsense syllables and relearning them after 20 minutes, 1 hour, 9 hours, and 1, 2, 6 or 31 days, the same intervals Ebbinghaus used. Memory was measured as “savings”: the share of the original learning time saved when relearning a list.

Time since learning Savings in the 2015 replication Savings in Ebbinghaus’s 1885 data
20 minutes 0.472 0.582
1 day 0.317 0.337
31 days 0.041 0.211

The curve matched Ebbinghaus’s closely, with a small upward jump at 24 hours. Treat it with care, though: this is one person learning meaningless syllables, and savings is not the share of a book you remember. That is why you will not find a “you forget 70% in a day” line here.

There is a second reason you forget: rereading feels like learning. In a 2006 study in Psychological Science, Roediger and Karpicke found that students who restudied a passage did better on a test 5 minutes later and felt more confident. Two days and one week later, though, the students who had tested themselves instead remembered substantially more. The comfortable method won the short race and lost the long one.

If you want the biology behind all this, Lisa Genova’s Remember explains how memories form, why we forget, and how to tell normal forgetting from the early signs of Alzheimer’s disease.

Test yourself instead of rereading

This is the habit with the clearest support. In Roediger and Karpicke’s experiments, students studied prose passages and then either restudied them or took one or three free-recall tests, writing down everything they remembered, with no feedback. On the final test 2 days or 1 week later, testing beat restudying. A 2013 review of ten study techniques by Dunlosky and colleagues in Psychological Science in the Public Interest rated practice testing as high utility, one of only two techniques to get that rating.

How to do it with a book: when you finish a chapter, close the book, then write or say the main points from memory. Only then open it again and check what you missed. The point is to pull the ideas out of memory before you look at them again.

The limits: the 2006 participants were college students reading short prose passages, we read that paper as an abstract, and no study here tested whole books.

Space out your reviews

The second high-utility technique is spacing. A 2006 meta-analysis in Psychological Bulletin by Cepeda and colleagues pooled 839 assessments from 317 experiments in 184 articles. Spaced study beat studying everything in one block, and the gap that worked best grew with how long people needed to remember.

A 2008 study in Psychological Science by the same group put numbers on that gap. More than 1,350 people learned facts, reviewed them once 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 the time until a test one week away, and about 5 to 10% of the time until a test one year away.

Here is our own arithmetic from those percentages, as an example rather than a finding: to remember something for a week, review it after about 1.5 to 3 days; to remember it for a year, review it after about 18 to 36 days. Dunlosky’s review also rates spaced practice as high utility.

The limits: these studies used word lists and facts, the 2008 study had a single review session, we read both as abstracts, and none of them used books. Benedict Carey’s How We Learn covers spacing and other study habits for a general reader, and it is part of The Fast Learner’s Foundation path.

What is the 2-7-30 rule?

The 2-7-30 rule says to review what you learned on day 2, day 7 and day 30. We searched research indexes on October 6, 2026 and found no study that tests this schedule.

The day counts look like the intervals Ebbinghaus used, which included 2, 6 and 31 days according to Murre and Dros. Ebbinghaus used them to measure forgetting, not to schedule reviews. Two studies help with the idea behind the rule, that the gaps should grow:

  • Karpicke and Roediger (2007), in the Journal of Experimental Psychology: Learning, Memory, and Cognition, ran three experiments with vocabulary word pairs. Growing gaps helped 10 minutes later, but equal gaps did better 2 days later. What mattered most was delaying the first test; growing the gaps had little effect on long-term retention.
  • Kang and colleagues (2014), in Psychonomic Bulletin & Review, taught foreign vocabulary over four weeks. Eight weeks later, growing and equal gaps produced the same recall, although the growing schedule kept recall higher during the training weeks.

Our verdict: 2-7-30 is a reasonable rule of thumb built on real findings, but it is untested as a schedule, and the best gap depends on how long you need to remember. Any spaced schedule with self-testing is better than rereading in one sitting. Keep in mind that these studies used word pairs and vocabulary.

Ask “why” and explain it in your own words

Dunlosky’s review gives a moderate rating to two related techniques: elaborative interrogation, which means asking why a statement is true, and self-explanation, which means explaining an idea to yourself as you go.

How to do it: after a key idea, stop and ask why it would be true. Then connect it to something you already know, or to something you have seen at work. If an idea in a management book reminds you of a meeting last week, that link is worth a sentence in your notes.

The limits: a moderate rating is weaker support than testing and spacing, the review was written for students, and its evidence runs up to about 2012.

Do handwritten notes beat a laptop?

You may have heard that taking notes by hand beats typing. The famous study is Mueller and Oppenheimer (2014) in Psychological Science: across three studies of lecture note-taking, laptop users did worse on conceptual questions, which the authors linked to copying lectures word for word. In a 2018 corrigendum, they kept the conclusions “qualitatively the same” but reported that the more informative effect sizes were “much smaller, ranging from .03 to .11”.

Later replications did not hold the finding up:

  • Morehead, Dunlosky and Rawson (2019), in Educational Psychology Review, repeated the study and added groups using an e-writer and taking no notes. They found no consistent differences between any groups, including the no-notes group, and called a verdict “premature”.
  • Urry and colleagues (2021), in Psychological Science, repeated the first study with 142 people (74 on laptops, 68 by hand). Laptop notes held more word-for-word text, but quiz scores did not differ, and a mini meta-analysis of eight similar studies agreed.

The takeaway: the tool matters less than what you do with the ideas. If you want your notes to help, write the ideas in your own words, which is the self-explanation habit from the previous section, not a lesson from these studies. All of them tested lecture notes and quizzes taken soon after, not notes on a book. For a full note-taking system, Sönke Ahrens’s How to Take Smart Notes explains the Zettelkasten method of connected notes.

What to stop relying on: highlighting and rereading

Dunlosky’s review rated highlighting, rereading, the keyword mnemonic and imagery for text as low utility. Rereading in particular is the trap from the first section: it feels like learning and raises confidence more than memory. Because of these low ratings, we also do not recommend mnemonic tricks such as the keyword method or mental images as a way to remember books.

The same review rated summarization as low utility too, but that means students writing their own summaries, not reading one. It may seem to clash with explaining ideas in your own words; the review rated the two techniques separately, and we report the ratings as given.

What about reading several books at once? A 2019 meta-analysis of interleaving (mixing topics instead of studying one at a time) by Brunmair and Richter in Psychological Bulletin pooled 59 studies. Interleaving helped overall (g = 0.42, a moderate effect), most for paintings (0.67) and less for math (0.34), but the evidence for expository texts was ambiguous, and for words, studying one block at a time did better (g = -0.39). The authors advise caution “especially for expository texts and words”. Interleaving research is about learning to tell categories apart, so we do not sell “read several books at once” as a memory technique.

Can you remember 90% of what you read?

People often ask how to retain 90% of what they learn, or how to train the brain to never forget. No study we reviewed supports a fixed percentage for either, and the forgetting-curve numbers above are savings scores for nonsense syllables, not percentages of a book.

Forgetting is normal, as the replication shows. The realistic goal is to keep the ideas you care about for longer, and testing and spacing are the two habits with the best support for that. Charan Ranganath’s Why We Remember explains how the brain chooses what to keep, which is a kinder way to think about the books you have half forgotten.

A simple routine for your next book

This routine adds no new claims; it puts the habits above in order.

  1. After each chapter, close the book and recall the main points without looking. Then check. (Testing.)
  2. Write two or three lines on why those points matter to you and what they connect to. (Self-explanation.)
  3. A few days later, recall the main points again before you look, then check them against your notes or a summary. (Spacing plus testing.)
  4. For ideas you want to keep for a year, do one more round a few weeks later; the arithmetic from the spacing studies points to roughly 18 to 36 days.

A summary has two honest uses in this routine. It works as the answer key after you test yourself, and as a short review of the main ideas a few days or weeks later. In a 1982 study in Memory & Cognition, Reder and Anderson found that spaced rereading of a text’s main points improved memory for its central ideas. That study used textbook passages, and we read it as an abstract.

Because most of this research used students, the routine fits exam reading directly. If you are building the habit, our list of books like Atomic Habits helps you make it stick, and the Learn Faster and Think More Clearly goals collect the books on learning. Wondering whether it matters if you listen instead of read? We looked at that research in audiobooks vs reading vs book summaries.

Studies cited

How BookBits fits

A BookBits audio summary works as the spaced review in step 3: a short pass over a book’s main ideas, written in our own words through an AI-augmented process and narrated with AI voices. It supports reading the book; it does not replace it. The four books mentioned on this page add up to about 1 hour 17 minutes of summaries. You can play a free preview on any book page without an account; full listening requires the iOS or Android app, Pricing covers the plan, and our page on the BookBits book summary app shows how it works.

Table of contents