Does Chess Make You Smarter? What the Research Actually Shows

No, not in the way most headlines claim. Chess reliably makes you better at chess. Whether it makes you smarter in the way people usually mean, meaning it raises your IQ or your grades at school, is a far weaker claim once you look at how the positive studies were actually designed, and the largest randomized trial on the question found no effect at all.

That does not make chess a waste of time, and it is not the conclusion most of the top ranking pages for this question reach. It is a distinction those pages mostly skip, and it matters if you are deciding whether to put a child, or yourself, into chess instruction expecting a cognitive payoff beyond the board.

What the positive studies actually found

The most cited positive evidence is Giovanni Sala and Fernand Gobet's 2016 meta analysis in Educational Research Review. It pooled 24 studies and 40 effect sizes and found chess instruction associated with an average gain of d=0.34 in overall cognitive ability, d=0.38 in math achievement and d=0.25 in literacy achievement among primary and middle school students. The effect grew with training hours, with somewhere around 25 to 30 hours, roughly a school year of weekly lessons, as the threshold where a meaningful effect appeared.

Those are real, published numbers, and they are the reason chess in schools programs can point to a body of research at all. What almost none of the roundups built on this meta analysis mention is a detail sitting inside the same paper: almost none of the 24 underlying studies used an active control group. Most compared a chess group against a group that received no additional activity at all, rather than against a group doing some other structured activity. That design cannot tell you whether chess itself caused the gain, or whether any novel, engaging activity taught by an attentive instructor would have produced a similar bump, the well documented placebo and expectation effect that shows up across educational research generally.

The same researchers walked it back

This is not an outside critique of Sala and Gobet's work. It is their own follow up. In a 2017 piece for The Conversation, the two researchers laid out what happens when chess trained groups are compared against active control groups, children doing checkers or sports instead of nothing, rather than passive ones: chess showed no significant cognitive advantage over the alternative activity. They placed that result alongside similarly weak transfer findings for music training and for working memory or brain training programs, three fields with the same pattern of early excitement not surviving a properly controlled comparison.

The largest study found nothing

The single most persuasive piece of evidence here is not a meta analysis of small studies, it is one large randomized trial. Jerrim, Macmillan, Micklewright, Sawtell and Wiggins ran a randomized controlled trial, funded by England's Education Endowment Foundation, that assigned about 4,000 pupils in roughly 100 English primary schools to either a year of chess instruction during the 2013/14 school year or to a control group that did not receive it. A year later, there was no evidence of improved math, reading or science attainment in the chess group, and no differential effect by gender or socioeconomic status. The trial is not beyond criticism: Sala and Gobet themselves argue in their 2017 piece that it has methodological shortcomings which probably invalidate its results, so it cannot prove that chess never helps. What it does show is that when chess instruction was rolled out across about 4,000 children, the gains the small studies reported did not reappear.

Put those side by side and the honest picture is this: the smaller studies without an active control lean positive, the comparisons against an active control show no chess advantage, and the largest trial found nothing. Effects that appear against doing nothing and vanish against doing something else are what a placebo or expectation effect looks like, and that is the most important fact for anyone deciding whether chess instruction will move a child's grades.

Smart people do play better chess, just not by as much as you would guess

There is a real correlation between general cognitive ability and chess skill, it is only smaller than the popular framing implies. Burgoyne, Sala, Gobet, Macnamara, Campitelli and Hambrick's 2016 meta analysis in the journal Intelligence, covering 82 effect sizes across 19 studies and 1,779 players, found a meta analytic average correlation of about r=0.24 between chess skill and general cognitive ability. Squared, that means cognitive ability statistically explains only around 6 percent of the variation in chess skill between players. The other 94 percent is left to everything else, practice and study among them.

That relationship is not constant across a player's career either. It is much stronger in young and unranked or novice players than in adult, ranked, skilled ones, and it is stronger for numerical ability than for verbal or visuospatial ability.

A Michigan State University press release on the same study frames the practical takeaway plainly: intelligence and chess skill are linked, but modestly, and the link matters far more for a young or novice player learning the game than for an elite adult competing at the top level.

Does chess cause intelligence, or is it the other way around

A correlation does not say which way the causation runs. Brighter children may be drawn to chess and do well early, or chess may sharpen them, or both. A study of young chess players by Bilalić, McLeod and Gobet adds a twist that matters for anyone training: in its small elite subgroup of 23 children, intelligence was not a significant factor in chess skill and, if anything, tended to correlate negatively with it. The explanation was practice. The less intelligent elite children practiced more, and practice is what got them to the same level.

Bilalić's broader body of research on chess expertise points the same direction from a different angle. It generally finds that chess experts outperform novices on tasks bound to the patterns of the game itself, but that this advantage is domain specific rather than evidence of generally elevated intelligence. Skill built at the board tends to stay at the board.

The small studies behind the popular narrative

The positive claims you see repeated online usually trace back to small, self selected studies rather than the large randomized trial. Kazemi and colleagues gave six months of chess instruction to 86 Iranian students across three grade levels and compared them to a 94-student control group, finding the chess group scored higher on posttest metacognition and math. Aciego and colleagues gave extracurricular chess instruction to 170 Spanish schoolchildren and reported a positive relationship with reasoning, creativity, hardiness and self confidence. Both are real, published studies, and both share the same limitation Sala and Gobet flag across the wider literature: no active control group, and in Aciego's case, children who chose to join a chess program rather than being randomly assigned to one, so the same children might well have shown the same gains from any comparably engaging activity. Neither study is dishonest or badly run for what it actually set out to test. The problem sits one step downstream, in how much later coverage borrowed the conclusion without carrying along the caveats that came attached to it.

What about dementia

Chess and dementia prevention is the one place where the evidence is genuinely interesting, with a real limit worth stating clearly. A 2019 scoping review by Lillo-Crespo and colleagues, covering 21 articles, concluded chess playing is associated with reduced dementia risk in people who do not yet have a diagnosis. It found no similar protective benefit in people who already have one, and the review itself flags a general shortage of strong scientific evidence on chess as a protective factor. This is an association built from observational data, people who choose to keep playing chess into old age may already differ from those who do not in ways that affect their dementia risk regardless of the chess itself, and it is worth remembering that the benefit only shows up before a diagnosis, not after.

Where the popular pages overstate it

Search this question and two of the top ranking pages illustrate the problem well. Chess.com's own roundup repeats a widely cited claim, attributed to Robert Ferguson, that a study of 4,000 Venezuelan students found significant IQ increases after four months of chess instruction. Neither that article nor any other page checked for this piece links to an accessible, peer reviewed publication describing that study's design or its control group, so treat the Venezuela claim as a popular story rather than verified evidence. The same Chess.com page also cites a 2017 Indian study on divergent thinking and a 2015 Iranian study on auditory memory in expert players, both small studies of the same kind Sala and Gobet warn against over generalizing from.

Brainscape Academy's own roundup goes further in the same direction. It presents only positive findings, the Venezuela claim, a dementia risk finding, and a brain imaging study, and no contradicting evidence anywhere in the piece. It never mentions the null result from the EEF randomized trial, and it never mentions that Sala and Gobet, the researchers behind the positive meta analysis it leans on, went on to raise the placebo effect problem themselves. That is not a small omission. It is the exact contradiction this page is built to correct.

So what actually helps your own game

Since the real skill here is domain specific rather than a general cognitive boost, the more useful question is not whether chess makes you smarter, it is what training actually transfers to your own games. If your puzzle rating and your game rating do not seem to agree, that gap has a real explanation, covered in why your puzzle rating doesn't match your game rating. If you want the broader answer on what actually raises your rating, hour for hour, we cover that directly in how to get better at chess, backed by data. If you have plateaued somewhere around eleven or twelve hundred and are wondering whether you have hit some kind of ceiling, you have not, and that is covered in why you stopped improving at 1200. If you are only just starting out, the rules and your first games are covered in how to play chess for beginners.

Chessdrive, the chess improvement app we build here, does not claim to make anyone smarter, and given everything above, that would be a strange thing to promise. What it is built for is the domain specific skill this page has been describing: a repertoire for both colors built from a short taste questionnaire, reviewed against a retention it computes from your rating and how often each position actually comes up. That is a narrower promise than a cognitive boost, and it is one we can actually back up.

The honest summary

Chess makes you better at chess. The evidence that it raises general intelligence or school attainment is real in smaller, less controlled studies, disappears against an active control group, and did not show up in the largest trial. The correlation between intelligence and chess skill is real but modest, and it shrinks the more serious a player becomes, which is the opposite of what a simple "chess equals higher IQ" story would predict. Among elite young players, practice mattered more than IQ. The dementia association is real and worth knowing, but it is observational, and it stops applying once a diagnosis is already made. None of that makes chess a waste of time. It makes chess exactly what it looks like: a demanding, rewarding game that builds a specific, transferable skill at chess itself, not a shortcut to a higher IQ.

Last verified: 28 September 2026, against the sources cited above.