Does highlighting actually help you study?
If you like highlighting, the research is uncomfortable reading.
In a well-known review of study techniques published in Psychological Science in the Public Interest, Dunlosky and colleagues assessed ten common strategies against the experimental evidence. Highlighting and underlining came out in the lowest tier — rated low utility, alongside re-reading and summarisation. Practice testing and distributed practice came out on top.
That finding is real and worth taking seriously. But the conclusion most people draw from it — “stop highlighting” — is the wrong one. What the studies actually show is narrower and more useful than that.
What the studies measured
The typical experiment hands students a passage, lets some highlight it and some not, then tests everyone later. Highlighting groups generally do no better than reading groups. On some question types they do slightly worse.
Note what that design captures: unguided highlighting, done on a first read, by people given no technique, followed by no further activity. That describes how highlighting is usually done — which is why the finding matters — but it does not test highlighting as a deliberate skill.
When researchers varied the conditions, the picture changed. Students who were trained to be selective, or who had to justify their marks, did better. The weak results track how the marking was done, not the act of marking.
Why it usually fails: the fluency illusion
Highlighting feels productive, and that feeling is the problem.
Recognising a sentence as important is easy and quick. It produces a pleasant sense of understanding — researchers call it fluency — that is only weakly related to whether you could reproduce the idea tomorrow. You end the session confident and under-prepared, which is worse than ending it knowing you are behind.
There is a second failure mode. Marking a sentence can function as deferral: you have flagged it as something to learn later, and the brain treats the flag as the work. The highlighter becomes a way of postponing thinking while feeling like you are doing it.
And volume kills it outright. Highlight 60% of a page and you have added no information, because the marks no longer distinguish anything. You have made the page harder to read and produced nothing to revise from.
Why selectivity is the whole ballgame
The one variable that consistently separates useful highlighting from useless highlighting is how much of the text gets marked — and not because sparse pages look tidier.
Deciding that this sentence matters more than that one requires holding both in mind, comparing them against what the chapter is arguing, and committing to a judgement. That is elaborative processing, and it is the kind of work that produces durable memory. It is the same reason explaining a concept to someone beats re-reading it.
Highlight everything and you skip the comparison entirely. Each mark becomes an isolated “yes” rather than a ranked judgement, and the cognitive work that would have made it stick never happens. The marks are a side effect of the thinking, not the point of it.
The conditions under which highlighting does work
Pulling the evidence together, highlighting earns its place when four things are true:
It is selective
Roughly one line in ten. Enough scarcity that every mark required a comparison against the lines you left unmarked.
It happens on a second pass
You cannot rank sentences by importance before you have read the argument they belong to. First read, then mark.
It is followed by retrieval practice
Convert each highlight into a question and answer it from memory. This is the step that moves you from low-utility to high-utility technique, and it is the step almost everyone omits.
It is a filter, not a study method
The purpose of highlighting is to reduce a chapter to the ten per cent worth testing yourself on. Producing that shortlist is the whole job. Re-reading it is not studying.
A workflow that respects the evidence
Read the section once, without marking anything. Go back and mark about a tenth of it, forcing yourself to choose. Write a question next to each mark. Close the book and answer the questions from memory. Check, and note what you missed. Come back to the ones you missed two days later, then a week after that.
Highlighting occupies one short step in that sequence. That is the correct amount of weight to give it — and it is also, usefully, the step that software can do for you. An automatic highlighter enforces the ten per cent budget and the second-pass logic by construction, which clears the way for you to spend your time on the questioning and retrieval that the research says actually moves the needle.