Does chess physically change the brain?

Learning can be accompanied by brain changes, but comparing experts and beginners is not enough to identify their cause. Observed differences must be distinguished from demonstrated effects of practice.

How long does it take to see an effect?

The sources presented do not establish a universal timescale. Comparing groups does not measure changes in the same brain during training.

Does chess protect against Alzheimer’s disease?

No preventive effect specific to chess is established here. You can play for enjoyment and social contact without treating it as a medical guarantee.

Do the benefits transfer to other skills?

Far transfer to school performance or general cognitive abilities is not firmly established. Findings depend in particular on the control groups used.

Chess and the Brain: What Studies of Experts Can Tell Us

A grandmaster recognises a position you still need to work through. It is tempting to imagine that a brain scan would immediately show what makes them strong. Research on players’ brains exists, but it cannot explain the origins of their expertise quite so simply.

Observing a difference and understanding its cause

A study may compare experts with beginners and find a difference in brain structure or activity. It is then measuring an association with expertise. The groups may also differ in their histories, selection, starting ages, environments or other activities.

Establishing an effect of training would require, among other things, following changes over time and using an appropriate comparison. A correlation between a brain measurement and years of practice is not enough, by itself, to distinguish all the possible explanations.

This does not make imaging useless. It clarifies the question it answers: “Do these groups differ on this measure?” is a different question from “Did chess produce this difference?”

The xiangqi case: check which game was studied

Li and colleagues (2015) published a multimodal MRI dataset involving 29 professional xiangqi players and 29 beginners. Xiangqi is the game often called Chinese chess; these participants should not be presented as a sample of experts in Western chess.

The work makes structural, diffusion and resting-state images available, alongside participant characteristics. It is a resource for studying expertise, not a trial in which beginners were randomly assigned to different training programmes.

Another precaution: several publications may use the same dataset. They do not then represent as many independent replications. Before counting “studies that confirm” a finding, check the participants, the game and whether data have been reused.

Executive functions: what Ramos actually shows

Ramos, Arán Filippetti and Krumm (2018) compare 65 children aged 8 to 12: 30 play chess and 35 have no experience of it. The authors report better performance in the chess group on the executive functions assessed.

The design is cross-sectional and ex post facto: the groups existed before measurement. This is neither a ten-week training intervention nor random assignment. We therefore cannot conclude that chess alone created the observed difference.

The reference year to retain is 2018, the year of the journal issue and publication indicated by the journal, even though the manuscript circulated in 2017. The coauthor’s name is Arán Filippetti, not “Ar”.

From improving at chess to improving elsewhere

You can learn to spot a fork without becoming better at every attention task. This distinction between specialised learning and far transfer is central to the review by Sala and Gobet (2017).

Their review examines findings from chess, music and working-memory training. Conclusions about transfer become less favourable when experiments better control expectations and the activities offered to other groups.

A group receiving chess lessons also benefits from an adult, a structured setting and a new activity. Comparing it only with a group receiving no additional activity leaves those explanations open. An active control helps distinguish them.

It would therefore be premature to promise a general increase in intelligence, better emotional control or a “stronger prefrontal cortex”. For links with school learning, see the article on chess and mathematics.

Ageing: an activity, not a guarantee

Playing can give you a stimulating activity and an opportunity to meet other people. That does not prove that chess prevents Alzheimer’s disease.

The WHO dementia fact sheet presents several modifiable risk factors and notes that there is no cure for dementia. It does not allow us to attribute a quantified reduction in an individual’s risk to chess. Confusing an association between leisure activities and health with a preventive prescription would be misleading.

How to read the next study

Before remembering a spectacular number, ask four questions: which game do the participants play? How were the groups formed? Does the measurement concern a task, a brain image or everyday life? Has the finding been replicated in other people?

For your training, keep observable goals: explaining a position better, remembering a pattern, checking a threat before moving. You can appreciate those improvements without making them carry a promise of general brain transformation.

Sources

Li, K., Jiang, J., Qiu, L. et al. (2015), A multimodal MRI dataset of professional chess players, Scientific Data, 2, 150044, article.
Ramos, L., Arán Filippetti, V. and Krumm, G. (2018), Funciones ejecutivas y práctica de ajedrez: un estudio en niños escolarizados, Psicogente, 21(39), 25–34, journal article.
Sala, G. and Gobet, F. (2017), Does Far Transfer Exist? Negative Evidence From Chess, Music, and Working Memory Training, Current Directions in Psychological Science, article.
World Health Organization (2025), Dementia, fact sheet, text.

Chess and the Brain: What Science Really Reveals