When people picture what a very intelligent person does with a free evening, they tend to imagine something difficult. A chessboard, maybe, or a puzzle of some kind. The assumption underneath is that intelligence works like a muscle and that hobbies are the reps.

It’s a comforting idea, partly because it implies intelligence is something a disciplined person can train directly. The research is less accommodating. The activity that shows up most reliably among people who score high on measures of intelligence isn’t demanding in that way at all, and it’s one most people quietly drop in adulthood.

It's reading. Not for a class or a job, but the ordinary kind nobody assigned.

Why Chess Can’t Do What This Hobby Does

Cognitive training research keeps running into a snag psychologists call the far transfer problem. Someone who practices a memory drill becomes very good at that drill, a little better at tasks that closely resemble it, and no better at much else. The improvement stays where it was made.

Chess is the honest illustration. Strong players hold an enormous amount of structured knowledge in their heads, and nearly all of it is knowledge about chess. Being able to see eight moves ahead on a board doesn’t always become the ability to see eight moves ahead in a hard conversation with a boss.

Reading behaves differently, and the reason has less to do with how much effort it takes than with what it deposits in a person over time.

What This Hobby Actually Builds

Psychologists separate fluid intelligence, the ability to reason through unfamiliar material, from crystallized intelligence, which is everything a person knows and can bring to a problem. Fluid ability is fairly stubborn and tends to peak young. Crystallized ability can keep growing for decades, and sustained reading is one of the most reliable ways of feeding it.

Raymond Cattell described this as investment. Fluid ability gets spent on particular domains across a lifetime, and whatever it’s spent on hardens into durable knowledge. Reading happens to be an unusually efficient place to spend it, because a book delivers language and ideas at a density ordinary conversation rarely reaches.

It also appears to matter what form the reading takes. A 2025 study published in Review of Educational Research , drawing on data from nearly half a million readers, found that leisure reading on screens has only a small association with comprehension, well short of the moderate link long established for print reading habits. The activity that predicts the outcome is closer to a book than to a feed.

The obvious objection is that intelligent people simply read more, which would make the hobby a symptom rather than a cause. That’s partly true, and conceding it makes the picture more useful rather than less.

A child who finds reading easy reads more, which builds the vocabulary and background knowledge that make the next book easier. The same loop keeps turning in adulthood, just more quietly and with nobody tracking it.

So the useful question isn’t whether reading makes anyone smarter in some clean causal sense. It's whether a person is inside that loop or outside it.

Plenty of what gets said about reading doesn’t survive scrutiny. Take the widely repeated finding that reading fiction sharpens theory of mind, meaning the ability to infer what other people are thinking and feeling. A 2024 study published in Journal of Experimental Psychology: General , pooling 70 experiments, found the effect is real but small, confined to empathy and mentalizing, and noticeably larger in published work than in unpublished work. That’s a long way from the version that circulated.

Reading also doesn’t function like a supplement. Nobody should expect a stack of novels to move a reasoning score. What the evidence supports is slower and less dramatic, which is a widening store of language and knowledge that makes new information easier to absorb.

It’s worth adding that the seriousness of the material matters less than people assume. Volume and variety do more work than any canonical list. And since a 2022 study published in Review of Educational Research found comprehension broadly comparable whether a text is read or listened to, audiobooks count, with the caveat that reading held a modest edge when a passage demanded real inference.

The Trait Underneath The Hobby

There’s a construct that helps explain why reading keeps appearing where other hobbies don’t. Psychologists call it need for cognition, a fairly stable tendency to seek out and enjoy effortful thinking. A 2023 study published in Proceedings of the National Academy of Sciences , synthesizing more than 1,300 studies, found that traits of this kind show larger relations with the knowledge-based side of intelligence than with raw reasoning. The disposition is less about horsepower than about how readily someone uses it when nothing obliges them to.

People high in this trait aren’t reading in order to improve themselves. They read because an unanswered question is mildly uncomfortable and a book relieves it. The hobby is what that disposition looks like when it has an evening free.

That changes what the question is really about. The goal was never to find the activity that trains the brain hardest. Intelligence, spread across a whole life, looks less like solving hard problems on command and more like continuing to take in new material long after anyone is checking.

The takeaway isn’t a page count or an ambitious reading list. It’s that the most reliable intellectual habit is also one of the least strenuous, which is probably why so many people let it go.

Curious which hobby your own mind reaches for when nothing is required of it? Find out how you actually take in and process the world with this science-inspired test: Cognitive Style Test