Rereading teaches you almost nothing: the method that works instead

You reread your course the night before the exam. Every page looks familiar. You close the folder feeling ready. The next morning, half the ideas are gone, and the rest come back blurry, without the words that went with them.

That gap is not mysterious. Rereading produces familiarity, and familiarity feels exactly like knowledge. You recognize the sentence, so you assume you know it. The trouble is that recognizing and recalling are two different operations, and only recall helps on the day nobody shows you the page: the exam, the meeting where someone asks for a figure, the bug you have to diagnose alone on a PC that has stopped responding.

This article explains why rereading misleads you, what works instead, and a fifteen-minute test you can run tomorrow on a single subject.

The feeling of knowing

When you reread a page, your brain recognizes it on the second pass. That recognition feels good, and it passes for understanding. You can follow an argument with your eyes without being able to rebuild it yourself, say it out loud, or answer a question someone asks you without the page in front of you.

Psychologists Jeffrey Karpicke and Henry Roediger made this gap measurable. Students who practiced recalling a text kept more of it over a delay than students who reread it or arranged it into concept maps. The result was published in Science in 2008: the effort of retrieval builds memory, while rereading only makes the material feel familiar.

The nuance matters. You do not need to reread less out of principle. You need to know what each block of study time is for.

The numbers that circulate

Two figures get repeated everywhere, and neither holds up.

The first says we only use 10% of our brain. Neurologist Barry Gordon, at Johns Hopkins School of Medicine, answers that we use virtually every part of the brain, and that most of it is active almost all the time. The myth flatters the idea that a magic setting could unlock the rest. There is no unused reserve waiting to be switched on.

The second is subtler. The “learning pyramid” assigns a precise retention rate to each way of taking in information: 10% for reading, 5% for a lecture, more for practice. Its origin is vague, and a 2020 analysis in Medical Education treats the pyramid as a myth that spread well beyond its evidence.

Why dwell on these figures? Because they reassure. They offer a simple rule (read less, explain more) without the mechanism that would justify it. A percentage says nothing about what you do with information after you have seen it. The useful criterion is not the channel (reading, listening, watching). It is the retrieval effort you make afterwards.

The gap that recall reveals

When you try to answer and nothing comes, you get the most useful signal of your whole week. The gap tells you exactly what to review, where, and how poorly you know it. Rereading never produces that signal, because every page looks in order. You spend your time confirming what you already know, and you carefully avoid what you do not.

That is also why a failed recall is not a failure. It is the data that tells you where to work. Three gaps found in ten minutes strike me as more useful than a two-hour reread that flags none.

Three shortcuts that feel like work

Highlighting. Running a yellow marker over a page creates the impression that you have processed the information. A few days later, you recognize the yellow, not the content. Highlighting marks what seemed important. It builds no memory.

The course open all the time. As long as the source sits next to you, you copy it instead of recalling it. Copying word for word is a polite form of rereading: you feel busy, and you only reproduce what your eyes see.

The three-hour session. After a while, the last half hour mostly serves to recognize pages you have already seen. You get tired, you click around, you feel virtuous. A shorter session with recall breaks builds more memory than a long reading marathon.

The fifteen-minute test

Run it on one subject: a chapter, a Windows procedure you keep forgetting, a concept from your job. Nothing more.

1. Open an empty text file and write the subject name at the top.
2. Close the source. For five minutes, write everything you remember: ideas, steps, examples. The gaps matter as much as what you get right.
3. Reopen the source and compare. Put a ? next to each point that is missing or wrong, and an OK next to those that held.
4. For each ?, reread only that passage, close the source again, then rewrite it from memory.
5. Schedule three follow-ups: tomorrow, in three days, in a week. A calendar event is enough.

The one-, three- and seven-day intervals are a practical rule, not a measured constant. What matters is coming back before the memory fades completely, and doing the recall again each time instead of rereading. By the second follow-up, your question marks should have dropped. If they stay the same, the problem is probably not memory. You likely do not understand the passage yet.

For a procedure, the test is even sharper. You write the steps in order without looking, and the steps you forget are the ones that will fail you when someone is waiting on you. A procedure you can only recognize is not a procedure you can perform.

Keep rereading for what it does well

Rereading is not useless everywhere. A quick pass through a chapter helps you map its structure: where the main parts sit and what depends on what. That map then gives your recall somewhere to hang. The trap starts when the map becomes the goal. You spend the week studying the map and never make the trip.

So: one pass to map, then recall to retain. That split keeps rereading for the job it does and stops it from doing every job.

When the test is not enough

The test works well for facts, definitions, procedures and vocabulary. For reasoning, it mostly tells you what you have forgotten, not whether you can argue a case or diagnose a problem. There, practice is the only real test: apply the idea to a concrete case and watch where it breaks.

And if your list of question marks contains ideas you do not understand at all, do not schedule three follow-ups right away. The test then produces only confusion. Go back to the explanation first, put it in your own words, and test afterwards. Recall is worth something only on an idea you already understand.

What this changes for you

Run the test tomorrow morning on the subject you really need to keep. Just one. In fifteen minutes it will tell you what your reading of the past week never could: what you actually know, and what never left the page.

For the bigger question, deciding what to keep and turning it into a system, the lesson Learning How to Learn sets out the decision framework, the place of recall in a real week, and the review tools. To connect this method to how you work, read what productivity really means too.

Sources

Figures and sources checked on 5 November 2026.

See the 3 sources for this article

• Jeffrey D. Karpicke and Henry L. Roediger III, 2008, The Critical Importance of Retrieval for Learning — Science
• Robynne Boyd, 2008, Do People Only Use 10 Percent of Their Brains? — Scientific American
• Ken Masters, 2020, Edgar Dale’s Pyramid of Learning in medical education: Further expansion of the myth — Medical Education