How to Take Notes From Books So You Remember What You Read

How to take notes from a book: a 2022 review of 36 articles found highlighting did little for understanding. Recall each chapter from memory instead.

An illustrated cover card headed “How to Take Notes From Books So You Remember What You Read”. Line drawing of a reading desk: on the left an open book with many lines highlighted in yellow; on the right the same kind of book lies closed beside a notepad filled with handwritten lines, a pen resting on it and a small tick in the corner.

You finish a business book on Sunday evening, its pages striped with yellow. On Wednesday a colleague asks what it said, and you can offer the title, one idea and a feeling that it was good. The highlights are still in the book. Very little of it came with you.

How to take notes from a book so you remember it comes down to one shift: spend less effort marking the page and more on producing something from it, above all writing down what you remember after each chapter with the book closed. In two experiments published in Science in 2011, students who read a science text and then recalled it from memory learned more than students who spent the same time building a concept map with the text open, though most did not expect it to.1 Six steps, best-tested first:

  1. Close the book after each chapter and write what you remember
  2. Check your page against the book, then recall again
  3. Come back to your recall page a few days later
  4. Explain the hard passages in the margin
  5. Add a question, a link or an example to each note
  6. Highlight little, and only after you understand the section

This guide stays with books. Lecture notes, and the old question of laptop versus pen, belong to why taking notes helps you learn, part of the site’s note-taking and knowledge management guides.

Left, a page striped with highlighter. Right, the book shut and a page written from memory.

Highlighting marks a passage but rarely teaches it

Highlighting on its own does little for understanding. Héctor Ponce, Richard Mayer and Ester Méndez pooled experiments from 36 published articles in a meta-analysismeta-analysis: A study that combines the results of earlier studies on the same question into one overall estimate. Pooling makes the estimate more precise, but it cannot repair the studies it pools: a meta-analysis of surveys is still survey evidence.Full entry in the glossary they published in 2022 and found that marking text as you read gave a small boost to memory for the text and no clear gain in comprehension. This guide relied on the paper’s abstract alone.2

The reason lies in what a highlight does. The 2013 review of ten study techniques led by Kent State University psychologist John Dunlosky explains that a marked line stands out the way one odd word does in a list, so it is easier to recall later. The review offers this as one possible reason highlighting helps with isolated facts more than with the links between ideas that understanding needs. In one study it describes, education students who underlined a long history chapter did worse on inference questions, a result the authors want replicated. Marking less keeps the effect alive: in another study, students allowed to underline only one sentence per paragraph recalled more of a science text than students who did not underline at all.3

Think of a manager reading a book on hiring who marks nearly every paragraph about structured interviews. The page looks studied. But choosing a line to mark took little thought, and nothing on the page records why structured interviews work or how she would run one.

Myth
A well-highlighted book is a well-learned book.
Fact
In a 2022 meta-analysis, highlighting gave a small memory benefit and no clear gain in understanding.

So treat a highlight as a bookmark for later work, not as the work. Lines exported from an e-reader are the same thing in a list: selections, not notes. Tiago Forte’s progressive summarization, which layers bold and highlighting over saved notes, has a different aim, finding a note quickly later, and the core ideas of building a second brain explain it.

Closing the book and recalling is the best-tested habit

Recalling what you read, with the book shut, has the strongest record of the habits in this guide. The 2013 Dunlosky review gave practice testing its top rating, high utility, and the meta-analyses summarized in a 2021 review by Garvin Brod of Frankfurt’s DIPF Leibniz Institute for Research and Information in Education found a moderate benefit of testing over restudying, in classrooms as well as labs.34

The study

Moderate evidence

Karpicke and Blunt's 2011 test of recall against concept mapping

Jeffrey Karpicke and Janell Blunt of Purdue University had 80 students study a science text once, four times, by building a concept map with the text open, or by recalling all they could, rereading and recalling again, with time matched for the last two. A week later the recall group did best, scoring about 50 percent higher than the mapping group on verbatim and inference questions combined. In a second experiment, 101 of 120 students did better on the text they had recalled than on the one they had mapped.1

Effort, in other words, is not all alike. Concept mapping is active work, yet mapping with the text open did less than the plainer act of recalling. The limits matter: short texts, undergraduates and a single test after a week, and a group of concept-mapping researchers challenged the study’s methods in a 2011 comment.5 Karpicke and Blunt also point out that a map drawn with the book closed turns into recall, and most students in their second experiment had predicted that mapping would serve them as well as recalling or better.1

Recall feels worse than it works, so expect the blank page after a chapter to be uncomfortable. Say you finish a chapter on negotiation and, book closed, manage three ideas and one story. Checking shows you forgot its main warning, about conceding too early. That gap is what you fix before the second recall, and the failed attempt still counted: Brod’s review notes that trying to retrieve something can help later learning even when the attempt fails, because it helps structure what you already know.4

Further reading

As an Amazon Associate WiserHours earns from qualifying purchases.

How to take notes from a book, chapter by chapter

The steps below put the best-tested habit first. They assume a nonfiction book read over a week or two, and each can shrink to a few minutes.

1. Close the book after each chapter and write what you remember

At the end of a chapter, shut the book and write the main ideas, the argument and any examples you can recall, in full sentences or a rough list. For a chapter on running meetings, that might be four lines: the claim, the evidence the author leans on, one example you liked and one thing you doubt. This is retrieval practiceretrieval practice: Learning by recalling material from memory, for example by quizzing yourself, instead of rereading it. Pulling information out makes it easier to recall later, a result also known as the testing effect.Full entry in the glossary in its simplest form, and the gaps you find become the to-do list for the next step.

2. Check your page against the book, then recall again

Open the chapter, compare, and fix what you missed or got wrong. Then close it and recall once more, the recall, reread, recall sequence Karpicke and Blunt used.1 Your corrected page is now your note for the chapter.

3. Come back to your recall page a few days later

Before you start a new chapter, or a few days after finishing the book, try to recall the page before rereading it. The Dunlosky review also gave its top rating to spreading study over time, called distributed practicedistributed practice: Spreading study or practice of the same material across several sessions with gaps between them, rather than cramming it into one. Also called spacing, it tends to give longer-lasting learning than the same time spent in a single block.Full entry in the glossary.3 Spacing and self-testing are the two habits the site’s advice on learning how to learn puts first, for any subject.

4. Explain the hard passages in the margin

When a passage surprises or confuses you, write one or two lines beside it on why it holds or what follows from it. Self-explanation costs time, so keep it for the passages that need it.6

Write the question the chapter leaves open, the earlier idea it connects to, or where it applies in your work. In the research Brod’s review summarizes, training in writing questions about a text consistently helped college students, and generating questions paid off when students had been taught how beforehand and given enough time.4

6. Highlight little, and only after you understand the section

Read a section through before marking it, then mark at most one sentence, the one you would quote. The Dunlosky review notes that most students mark too much, and a page that is mostly marked has nothing left to stand out.3

For each chapter you read

Good book notes add something the page does not say

Notes help you learn from a book when they make you produce something the text does not already contain, which researchers call generative learning. Logan Fiorella and Richard Mayer define it in a 2016 review as making sense of new material by reorganizing it and linking it with what you already know, and they describe eight strategies, from summarizing and self-testing to explaining and teaching.7

Brod’s review draws the line sharply. An activity counts as generative only when learners add meaningful content beyond what they were given, which rules out highlighting and simple paraphrasing. All six strategies he reviewed (concept mapping, explaining, predicting, questioning, practice testing and drawing) had worked for university students, while the evidence for younger students was mixed. The reason he gives is plain: to add content, you have to call up what you already know and tie the new material to it.4

Take a hypothetical chapter on pricing arguing that buyers anchor on the first number they see. Copying that sentence adds nothing. Writing “our renewal quotes open with the discount; try opening with the full price” adds an example, a link to your own work and something to test. That is the test for any book note: what does it say that the page does not?

Handing the work to a chatbot has been tested against notes, with young readers. In a trial in English schools published online in 2025 by researchers at Cambridge University Press & Assessment and Microsoft Research, a company that invests in generative AI, 14- and 15-year-olds understood and remembered a history passage better three days later when they had taken their own notes than when they had only used a chatbot.8

Margin notes work best as explanations, not copies

Explaining a passage to yourself as you read, called self-explanation, had a moderate average benefit across 64 research reports in a 2018 meta-analysis by Kiran Bisra and colleagues at Simon Fraser University in Canada. Studies in which people learned from texts showed one of the larger averages, though the differences between types of task were not statistically clear.6

Why would explaining help? Bisra’s team describes self-explanation as a way of inferring the causes and relationships a text leaves unstated.6 Reading “retention fell after the reorganization” is passive; writing “because teams lost their manager mid-project?” in the margin is an inference you now have to check against the next page.

The catch is time. The Dunlosky review, which rated self-explanation moderate utility, describes a classic 1994 study in which eighth graders who explained each sentence of a text on the circulatory system learned more than classmates who read each sentence twice, but spent about twice as long.3 In the five studies in Bisra’s analysis that kept study time equal, the benefit was still clear, though smaller on average, a gap that was not itself statistically clear.6

No controlled study of readers’ own margin notes as such appeared in our searches; the margin is simply a handy place for the explanations and questions this research does support. So spend explanations where a passage is hard or surprising, not on every line. Asking why the idea should be true, or how you would explain it to a newcomer as the Feynman technique suggests, gives the note something to say.

Mark, explain, recall: each asks more of you than the one before, and the last is the best tested.

Which reading habits the evidence backs, and how firmly

For remembering what you read, recall from memory has the strongest evidence, explaining passages to yourself moderate evidence, and highlighting the weakest, according to the reviews and meta-analyses cited here. Nearly all of that evidence was gathered with students reading short texts.

What it is What the best evidence found Evidence
Highlighting while reading Small benefit for memory (average effect size 0.36), no clear gain in comprehension; helped college but not school students Meta-analysis of 36 articles, limited; abstract only2
Recalling after reading Practice testing rated high utility; testing beat restudying by a moderate margin across meta-analyses Expert review and meta-analyses, strong for students34
Explaining to yourself Average benefit g = 0.55 across 64 reports; g = 0.41 where study time was equal Meta-analysis, moderate6
Notes that add content All six generative strategies reviewed worked for university students Narrative review, moderate4
Own notes vs a chatbot Own notes beat chatbot use alone on retention and comprehension 3 days later (ages 14 to 15) One preregistered trial, moderate8
Rereading Rated low utility Expert review, limited3

The bottom line

What you do after a chapter matters more than what you mark during it. Close the book, write what you remember, check it and recall again, then add the explanations, questions and examples the page does not supply. Highlight if you like, but sparingly, and treat it as the start of your notes rather than the notes themselves.

Frequently asked questions

Can an AI summary of a book replace your own notes?

Not if you want to remember the book, going by a trial in English schools published online in 2025 by researchers at Cambridge University Press & Assessment and Microsoft Research. Students aged 14 and 15 who took their own notes on a history passage did better three days later than when they only used a chatbot, although most preferred the chatbot. The study used short passages, not books, and adults were not tested.

Are concept maps good book notes?

They can be, if you draw them from memory. A 2018 meta-analysis summarized in Garvin Brod's 2021 review found that making your own concept maps helped students learn. But in Jeffrey Karpicke and Janell Blunt's 2011 experiments, mapping with the text open did worse than recalling it, and the two researchers point out that a map drawn with the book closed becomes a form of recall.

Is rereading a book a good way to remember it?

It is one of the weaker options. The 2013 review of ten study techniques led by Kent State University psychologist John Dunlosky rated rereading low utility: its benefits were shown with many kinds of text but not well established for other learners, tasks and settings. Practice testing got that review's top rating, so recall first and reread only to check what you missed.

Sources

  1. Retrieval Practice Produces More Learning than Elaborative Studying with Concept Mapping. Karpicke, J. D. & Blunt, J. R. (2011). Science, 331(6018)
  2. Effects of Learner-Generated Highlighting and Instructor-Provided Highlighting on Learning from Text: A Meta-Analysis. Ponce, H. R., Mayer, R. E. & Méndez, E. E. (2022). Educational Psychology Review, 34(2)
  3. Improving Students' Learning With Effective Learning Techniques: Promising Directions From Cognitive and Educational Psychology. Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J. & Willingham, D. T. (2013). Psychological Science in the Public Interest, 14(1)
  4. Generative Learning: Which Strategies for What Age? Brod, G. (2021). Educational Psychology Review, 33(4)
  5. Comment on “Retrieval Practice Produces More Learning than Elaborative Studying with Concept Mapping”. Mintzes, J. J., Canas, A., Coffey, J., et al. (2011). Science, 334(6055)
  6. Inducing Self-Explanation: a Meta-Analysis. Bisra, K., Liu, Q., Nesbit, J. C., Salimi, F. & Winne, P. H. (2018). Educational Psychology Review, 30(3)
  7. Eight Ways to Promote Generative Learning. Fiorella, L. & Mayer, R. E. (2016). Educational Psychology Review, 28(4)
  8. Effects of LLM use and note-taking on reading comprehension and memory: A randomised experiment in secondary schools. Kreijkes, P., Kewenig, V., Kuvalja, M., et al. (2026). Computers & Education, 243, 105514; published online 2025

How we researched this

Searches of PubMed, ERIC, OpenAlex, Crossref and general web search in September 2026 looked for reviews, meta-analyses and experiments on learning from text: highlighting, generative learning, self-explanation and retrieval practice. The sources run from 2011 to 2026; all but three were read in full, and those three as abstracts. The chief limitation: nearly every study tested students on short texts over days or weeks, not adults reading whole books.

Last updated . Read our editorial policy.

Cite this article: WiserHours. (2026). How to Take Notes From Books So You Remember What You Read. WiserHours. https://wiserhours.com/note-taking/notes-from-books/. Tables and charts may be reused with a link back to this page.