To remember what you read, stop rereading and start recalling. Close the book or the tab and try to say back what it told you, from memory. That act of retrieval does more for memory than rereading or highlighting. Read with a question in mind, mark very little, write three sentences from memory when you finish, revisit those sentences a few days and a few weeks later, and tie each idea to something you already know.
That's the short version. The rest of this post is why it works, and how to fit it into the reading you actually do, which for most of us is half books and half things on a phone at 11pm.
I'll admit why I care. I build an app for saving things, and for years I believed that saving something was most of the way to knowing it. It isn't. It's barely the first step.
Why you forget most of what you read
The forgetting you notice isn't a personal failing. It's the default.
In the 1880s Hermann Ebbinghaus memorized lists of nonsense syllables, then measured how much easier they were to relearn after different delays. Memory dropped steeply in the first hours and days, then flattened out. In 2015, researchers at the University of Amsterdam repeated the experiment and got a curve that looks much the same, across delays from 20 minutes to 31 days.
Real reading isn't nonsense syllables. Meaningful material sticks better than "WID" and "ZOF". But the shape holds. Most of the loss happens early, which means what you do in the hour and the week after reading matters far more than how carefully you read the first time.
There's a second problem. Rereading feels like learning. The text gets familiar, the familiarity feels like knowledge, and you close the book confident. Then a month later you can't name the three main arguments.
What actually works, according to the research
Retrieval beats rereading
The clearest evidence comes from a 2006 study by Henry Roediger and Jeffrey Karpicke. Students read short prose passages. Some reread them; others took a recall test where they wrote down everything they could remember, with no feedback.
Five minutes later, the rereaders were slightly ahead, 81% to 75%. One week later, it flipped: the students who had practiced recalling remembered 56% of the material, and the rereaders 42%. In a second experiment, students who studied once and recalled three times remembered 61% after a week, against 40% for those who studied four times. The repeat-readers were also the most confident they would remember. They were wrong.
This is called the testing effect, or retrieval practice. Pulling an idea out of your head strengthens the path back to it. Putting it in again, by rereading, mostly doesn't.
Spacing beats cramming
A big 2013 review led by John Dunlosky rated ten common study techniques. Only two earned the top rating: practice testing and spreading practice out over time. Highlighting, rereading and summarizing landed in the low-utility group.
That last one needs a footnote, because I'm about to tell you to write a summary. The review found summarizing hard to do well and unreliable as most students practice it, with the text open in front of them, copying it down in shorter form. Writing three sentences with the book closed is a different exercise. It's retrieval wearing a summary's clothes.
Connecting beats collecting
New information sticks when it hooks onto something you already know. Asking yourself "why would that be true?" or "where have I seen this before?" forces that connection. Researchers call it elaboration. You can just call it thinking about what you read.
How to remember what you read: a five-step method
None of this needs a system, an app or a notebook with tabs. It needs about five extra minutes per book chapter or long article.
1. Read with a question
Before you start, write one question you want the piece to answer. "Why did the Roman Republic fall?" "What should I actually change about my sleep?" A question gives your attention somewhere to land. You'll notice the answer when it arrives, and you'll remember noticing it.
2. Mark sparingly
Allow yourself one mark per page or section, at most. Highlighting everything is a way of postponing the decision about what matters, and the research above suggests it buys you very little. If you can only mark one sentence, you have to judge which one. That judgment is the useful part.
3. Close it and write three sentences
When you finish a chapter or article, close it. Then write three sentences from memory: what it said, why it matters, and what you'd do differently. Don't peek until you're done. Then check what you missed, and notice it. The gaps are exactly where the next review should go.
4. Come back at spaced intervals
Look at your three sentences again the next day, then about a week later, then about a month later. Don't reread them first. Try to recall them, then check. Each review takes a minute, and each one resets the forgetting curve to a gentler slope.
A calendar reminder works. So does a sticky note inside the cover of the next book you read.
5. Connect it to one thing you already know
Finish with one link: an idea from another book, something from work, an argument you had last week. "This is the same as..." or "this contradicts..." is enough. Ideas with connections are easier to find in your own head later, for the same reason a well-linked page is easier to find on the web.
How to remember what you read in a book
Books are long, so the method needs a little scaffolding.
- One card per chapter. After each chapter, write your three sentences on an index card or a note on your phone. A 15-chapter book becomes 15 short cards, which is a review you can do on a train.
- The one-page book. When you finish, write a single page from memory: the main argument, the three best ideas, and one thing you disagree with. Then compare it against your chapter cards.
- Tell someone. Explaining a book to a friend over dinner is retrieval practice with an audience. If you can't explain it, you know which chapter to revisit.
- Reread selectively, months later. Rereading isn't useless. It's just a poor first move. A second read of your marked passages six months later, after you've tried to recall them, is a good use of an evening.
Articles, threads and screenshots: the reading you do on your phone
Most of what we read isn't books. It's a long article on the bus, a thread someone shared, a page from an ebook you screenshotted because one paragraph hit hard.
The same rules apply in miniature. After a good article, take thirty seconds and say its point to yourself in one sentence. If something is worth keeping, keep it with a line of your own words attached: "why I saved this". That line is a tiny act of retrieval, and later it tells you why you cared.
The trap is that phone reading makes saving effortless, so we save instead of thinking. Which brings me to the uncomfortable part.
Saving isn't the same as remembering
Bookmarking an article feels like progress. So does a camera roll full of screenshotted book pages, or a read-later queue with 600 items. But nothing in your head changed when you tapped save.
There's a name for this trap, the collector's fallacy: the feeling that collecting information is the same as learning it. I've written separately about why read-it-later queues only ever grow. The short version is that a saved item is a promise, and promises without a review step don't get kept.
So be honest about the two jobs. Memory is for the handful of ideas you want to carry around and use without looking. An external store is for everything else: the quote, the statistic, the recipe, the exact page. You won't remember all of it, and you don't need to, as long as you can find it again.
Where Reminari fits
Reminari is built for that second job. You save from the share sheet: an article, a link, or a screenshot of the page you just read. It reads the text inside screenshots on your device, writes a short summary with key points, and lets you search by meaning instead of exact words. Months later you can ask "what was that book passage about attention residue?" and get an answer from your own saves, with the sources shown.
It won't do your remembering for you, and it shouldn't claim to. Use the five steps for the ideas you want in your head, and let something searchable hold the rest. Reminari is Android-first, with iOS in active development; you can join the early access list here.
Frequently asked questions
How can I remember what I read in a book?
Write three sentences from memory after each chapter, without looking back, then check what you missed. Review those notes the next day, a week later and a month later. Finish the book by writing a one-page summary from memory and explaining it to someone.
Why do I forget what I read so quickly?
Forgetting is steepest in the first hours and days after reading, a pattern first measured by Hermann Ebbinghaus and replicated in 2015. Rereading also creates a feeling of familiarity that's easy to mistake for knowing. Recalling the material soon after reading slows that early loss.
Is highlighting a waste of time?
Mostly, if it's all you do. A 2013 review of study techniques rated highlighting as low utility because it rarely improves later recall. Marking one key sentence per page is fine; the value is in deciding what matters, not in the color.
How often should I review what I read?
A simple schedule works: the next day, about a week later and about a month later. Each time, try to recall the main points before you look at your notes. Spacing reviews out like this is one of the two techniques with the strongest evidence behind it.
Should I take notes while reading or after?
After is better for memory. Taking notes while reading tends to become copying, while writing from memory once you've finished forces retrieval. Keep in-text marking light, then do the real note-taking with the book closed.
