Key Findings:
- A 2025 study of 400 kindergarten children found statistically significant math improvements in chess-playing groups compared to non-chess controls (p < 0.001), with WIAT-III mathematics scores of 85 and 84 versus lower control group scores.
- A 2017 Danish study found that replacing one weekly math class with chess improved standardized math scores by 0.16 to 0.18 standard deviations in grades 1 to 3, without harming overall math outcomes.
- The largest randomized controlled trial on chess and school performance, involving 4,009 UK pupils across 100 schools, found no evidence that chess improved math, reading, or science scores compared to control groups (Jerrim et al., 2018).
- A 2016 meta-analysis of 24 studies initially reported an average math improvement of +0.34 standard deviations from chess instruction. When studies with active control groups were isolated, this effect dropped to near zero.
- The key dividing line in the research is not whether chess helps, but under what conditions. Structured chess integrated into the curriculum with certified chess instructors shows stronger results than chess as an after-school club.
- Children with learning disabilities showed specific improvements in basic addition and counting after one year of chess instruction, though no improvement was found in complex multi-step calculation (Scholz et al., 2008).
Why This Question Matters to Parents
Parents searching for the right activity for their child often land on the same claim. Chess makes children better at maths.
It sounds logical. Chess is strategic. Maths is logical. Both require thinking ahead. The connection feels obvious.
But the research tells a more complicated story. Some studies show clear improvement. Others find no effect at all. The difference between those two outcomes comes down to how chess was taught, who was studied, and how the math was measured.
This report goes through all of it. We looked at the direct intervention studies, the meta-analyses, the failed RCTs, the school programs, and the methodological problems that have clouded this debate for thirty years. Where the data is clear, we say so. Where verified evidence does not exist, we flag it plainly.
Direct Studies on Chess and Math Performance
Scholz et al. (2008): Chess and Children with Learning Disabilities
Sample: 53 primary school children with learning disabilities, IQ scores between 70 and 85, drawn from seven classes across four specialized schools in Saxony, Germany.
Method: Two groups. The experimental group of 31 students received one hour of chess per week integrated into mathematics resource classes. The control group of 22 students received standard math instruction only. Duration: one full academic year. ANCOVA model used to adjust for pre-test baseline differences.
| Outcome Measured | Chess Group | Control Group |
|---|---|---|
| Basic addition and counting | Significantly higher | Lower |
| Complex written calculation | No significant difference | No significant difference |
| Gap tasks | No significant difference | No significant difference |
| General concentration | No significant difference | No significant difference |
Source: Scholz, M. et al. (2008). Impact of Chess Training on Mathematics Performance and Concentration Ability of Children with Learning Disabilities. Heilpadagogische Forschung.
Conclusion: Chess helped with foundational arithmetic. It did not improve complex, multi-step mathematical problem-solving. Scholz et al. concluded chess is a valuable educational tool for children with learning disabilities when expectations focus on fundamental numeric operations.
Ferguson (1995): Aggregated US and International Data
| Study | Sample | Grade Level | Key Math Finding |
|---|---|---|---|
| New Brunswick (Gaudreau, 1992) | 437 students | Grade 5 | Chess group outperformed control by 21.46% in problem-solving; pre-test 62% to post-test 81.2% |
| Texas Study (Liptrap, 1994-1997) | Elementary, 4 schools | Grades 3 and 5 | Chess club participants scored 6.4 TLI points higher in math (p < 0.00001); twice the improvement of non-participants |
| Zaire Study (Frank, 1970s) | 92 students | Ages 16 to 18 | Chess group showed significant advancement in numerical aptitude and spatial reasoning |
Source: Ferguson, R.C. (1995). Chess in Education Research Summary. American Chess Foundation.
The Liptrap Texas study has since been heavily criticized for self-selection bias. Chess club participants were 3:1 male and overrepresented by gifted-and-talented students. High-ability math thinkers likely chose chess rather than chess creating high-ability math thinkers.
Trinchero (2013): Italian Primary School Intervention
Sample: 568 primary school pupils aged 7 to 11, drawn from schools in the provinces of Asti and Bergamo, Italy.
Method: Four-group randomized design controlling for test sensitization. Math outcomes measured using the OECD-PISA Mathematics Scale. Instruction combined in-person certified chess instructors with online training modules.
Key finding: Small but statistically significant increase in problem-solving on complex math tasks. A dosage effect was identified: students who attended more chess hours and progressed further online showed proportionally greater math gains.
Conclusion: Chess supports acquisition of mathematical abilities when used alongside formal instruction, not as a replacement, and when designed to develop habits of mind including self-reflection and strategic perseverance.
Ye (2025): The Strongest Recent Data
Sample: 400 kindergarten children aged 5 to 6. Split into two educational models against non-chess control groups.
| Group | WIAT-III Math Score (Post-Test) | vs Control (p-value) |
|---|---|---|
| Classroom Integration Model (n=200) | 85 | p < 0.001 (t = 4.51) |
| Extracurricular Model (n=200) | 84 | p < 0.001 |
| Non-Chess Control Group | Lower | Baseline |
Source: Ye, Y. (2025). Frontiers in Psychology. DOI: 10.3389/fpsyg.2025.1592247.
Both chess groups showed highly statistically significant improvements over the control group. The classroom integration model showed a slight consistent advantage across all cognitive metrics including memory, attention, reading, and mathematics.
Rosholm et al. (2017): The Danish Curriculum Study
Sample: 482 students in grades 1 to 3 in Denmark.
Method: One weekly math class was replaced with chess instruction. Math outcomes measured using standardized tests.
Key finding: Replacing one math class per week with chess led to a 0.16 to 0.18 standard deviation improvement in standardized math scores, without harming overall math outcomes despite spending less time on direct math instruction.
Source: Rosholm, M. et al. (2017). Your Move: The Effect of Chess on Mathematics Test Scores. PLOS ONE.
Why Chess and Math Are Connected
Four overlapping brain systems link chess play to mathematical thinking. The research on these mechanisms helps explain why some studies find positive results.
- Visuospatial Skills and Spatial Reasoning: Chess requires continuous mental manipulation of geometric patterns across an 8×8 grid. Yakushina et al. (2025) confirmed that chess relies heavily on visuospatial working memory. Gao et al. (2019) found that children who played chess daily for five years showed superior visuospatial perspective-taking compared to non-players. This directly applies to geometry.
- Logical Reasoning and Pattern Recognition: Ye (2025) highlights that strategic planning in chess involves logical-mathematical intelligence and intense pattern recognition. Chess experts recognize familiar structural patterns rather than calculating every move, a skill known as chunking. This transfers to identifying mathematical sequences and solving algebraic equations.
- Working Memory: Estrada-Plana et al. (2024) note that storing, updating, and manipulating information in real time during chess actively strengthens working memory. A well-developed working memory, trained by tracking piece values and anticipating opponent responses, allows a student to process mental arithmetic without losing track of operative steps.
- Executive Function and Inhibitory Control: Estrada-Plana et al. (2024) identify that the rule-based, predictable nature of chess develops inhibitory control, the ability to suppress impulsive actions. In mathematics, inhibitory control is what prevents a student from guessing an answer rather than working through a complex formula step by step.
What Neuroscience Shows
A 2019 review by Ortiz-Pulido et al. in the Revista Mexicana de Neurociencia analyzed fMRI data comparing novice, intermediate, and advanced chess players. The study confirmed that playing chess recruits cortical regions responsible for creativity, anticipation, long-term planning, and memory. These brain regions directly overlap with those required for complex mathematical reasoning and executive function.
Chess in School Programs and Math Outcomes
Chess in Schools and Communities (CSC), United Kingdom
The CSC program, founded by Malcolm Pein, is a national UK charity integrating chess into state primary education.
The EEF Randomized Controlled Trial (Jerrim et al., 2018): 4,009 pupils tracked across 100 primary schools in 11 Local Education Authorities over the 2013/2014 academic year. Finding: no evidence that chess instruction had a positive impact on mathematics, reading, or science test scores compared to the control group. No positive academic impact for children eligible for free school meals either.
Despite null quantitative findings, Ofsted recognized partner schools involved in the CSC program for improved overall school quality, crediting enrichment partnerships including chess.
FIDE Chess in Education: Global Program Data
Armenia (2011): Chess made compulsory for grades 2 to 4 nationally. Mixed math outcome data from the ISET Policy Institute. For girls, improvement in chess showed a small positive link to math. For boys, no statistically significant connection was found. General IQ scores occasionally showed stagnant or declining trends in grade 4.
Russia: An 18-year study rooted in Lev Vygotsky’s developmental psychology framework demonstrated long-lasting positive effects of chess on school careers, particularly for disadvantaged students.
USA: Generally positive math outcomes from early studies. Principal Cheryl Coles of Public School 68 in the Bronx reported that standardized math scores increased school-wide by 18.6% following chess program implementation (California Achievement Test, 1995).
Expert Opinions and Meta-Analyses
Sala and Gobet (2016) Meta-Analysis
Published in: Educational Research Review.
Studies reviewed: 24 empirical studies measuring the relationship between chess instruction and children’s academic outcomes.
Initial finding: The raw aggregation suggested an average effect size of g = +0.34 standard deviations for mathematics improvement.
Critical finding: When studies with active control groups were isolated, the far-transfer benefits of chess on mathematics dropped to near zero.
Conclusion: While near transfer (improving at spatial tasks similar to chess) is reliable, the overall effect of far transfer to generalized mathematics is null or highly modest. Early studies likely suffered from severe placebo effects caused by the novelty and excitement of chess as an activity.
Expert Views
Dr. Giovanni Sala (Fujita Health University): The overall effect of far transfer is null, and there is no evidence that cognitive training programs improve academic performance. Professional and educational curricula should focus on domain-specific knowledge rather than general and allegedly transferable skills.
Malcolm Pein (CEO, Chess in Schools and Communities, UK): Children taught chess showed improved concentration, numeracy, logical thinking, pattern recognition and problem solving. It teaches them how to lose and how to win gracefully.
FIDE Official Position
FIDE’s Chess in Education Commission states that the 8×8 grid relies on geometric principles, combinatorics, the Shannon number (the mathematical calculation of possible game complexities), and probabilistic analysis. FIDE’s official position is that analyzing decision trees and practicing chess puzzles builds problem-solving abilities that mirror the skills required to solve mathematical equations and logical proofs.
FIDE launched the STEM CHESS 2025 project in collaboration with Opening Master, calling upon STEM graduate students to conduct research using a database of over 10.3 million chess games.
Where the Evidence Is Weak or Mixed
Studies That Found No Math Improvement
Jerrim et al. (2018): The largest and most rigorous RCT in this field. 4,009 pupils, 100 UK schools, funded by the Education Endowment Foundation. Conclusion: no evidence of any effect of chess instruction on mathematics, reading, or science scores. Jerrim warned that previous literature likely contains numerous false-positive effects driven by less robust research designs and small sample sizes.
Sala and Gobet (2017): Two experimental studies published in Learning and Behavior. When the control group engaged in an equally stimulating alternative activity, chess-trained children did not significantly outperform controls in mathematical problem-solving. Conclusion: the special academic effect of chess largely evaporates when the placebo effect is properly removed.
Methodological Criticisms of Early Research
| Criticism | Studies Affected | Details |
|---|---|---|
| Self-selection bias | Liptrap (1994-1997) | Chess club participants were 3:1 male and overrepresented by gifted students. High-ability thinkers chose chess, not the reverse. |
| Lack of active control groups | Most pre-2010 studies | Control groups did nothing novel. Chess groups benefited from Hawthorne effect (novelty-driven attention boost). |
| Small sample sizes | Ferguson (1995) primary study | Some foundational studies used as few as 15 students in the treatment group. Ferguson himself acknowledged this limitation. |
| Replicability problems | Most early US studies | Many findings have not been successfully replicated under controlled conditions in the two decades since. |
Source: Sala and Gobet (2016), Educational Research Review; Jerrim et al. (2018), Education Endowment Foundation.
Supporting Statistics
Global Chess Participation in Schools
According to the FIDE-ECU Chess in Schools Survey Report, approximately 25.6 million children participate in chess activities within a school environment globally. This is supported by approximately 92,550 recognized school teachers and 145,690 external chess coaches worldwide. The majority are in Asia, with an estimated 15 million in India and 5 million in China.
Average Math Improvement Across Study Types
| Study Type | Control Method | Average Reported Math Improvement |
|---|---|---|
| Meta-analysis average (Sala and Gobet, 2016) | Passive / uncontrolled | +0.34 standard deviations |
| Curriculum integration (Rosholm et al., 2017) | Active, Denmark | +0.16 to +0.18 standard deviations |
| Randomized controlled trial (Jerrim et al., 2018) | Active control, UK | 0.00 (null effect) |
Chess, Girls, and the Math Gender Gap
Globally, females represent approximately 11% of mixed-sex tournament chess players and just 2% of Grandmasters. Research by Cubel and Sanchez-Pages found that women and men of identical Elo ratings perform similarly in single-gender games, but women statistically underperform in mixed-gender settings. This is largely attributed to stereotype threat and risk aversion.
Data from the ISET Policy Institute tracking Armenia’s mandatory chess program found that girls had a higher baseline in mathematics than boys but lower average chess performance. Longitudinal data showed a small positive link between chess improvement and math performance for girls. For boys, no such causal connection could be statistically established.
DATA GAP: No large-scale, gender-controlled RCT on chess and math outcomes exists as of April 2026. The Armenian data is observational and context-specific.
What This Means for Parents:
Chess does not automatically make children better at maths. The largest and most rigorous study found no effect. If someone tells you chess will improve your child’s math scores, the honest answer is: it depends entirely on how chess is taught.
Chess that is structured, regular, and integrated with the curriculum shows real promise. The Ye (2025) study, the Trinchero (2013) study, and the Rosholm (2017) Danish study all show meaningful gains, but only when chess instruction was deliberate, frequent, and connected to learning goals.
The cognitive skills that chess genuinely builds are real and well-documented. Visuospatial memory, pattern recognition, working memory, inhibitory control, and logical reasoning all show measurable improvement in chess-playing children. These are the same cognitive tools that underpin mathematical thinking.
The quality of instruction matters more than the activity itself. A child learning chess from a structured curriculum with a qualified coach will develop more transferable cognitive skills than a child playing unguided games on an app.
Research Gaps and Limitations
This report reflects verified research available as of April 2026. The following gaps are flagged honestly.
DATA GAP: No large-scale RCT has been conducted specifically comparing chess and math outcomes for children aged 5 to 9 in the United States using a fully active control group.
DATA GAP: The Bart (2014) study cited in multiple chess education texts could not be independently verified for its specific empirical math outcome data.
DATA GAP: No long-term longitudinal study tracks whether early chess-related math gains persist into secondary school and beyond.
DATA GAP: No verified data exists on whether online chess instruction specifically produces equivalent math cognitive transfer effects compared to in-person instruction.
DATA GAP: No large-scale, gender-controlled RCT on chess and mathematics outcomes has been conducted as of April 2026.
DATA GAP: The cumulative number of children reached by Chess in the Schools NYC could not be independently verified from current public data.
Full Reference List
- Scholz, M. et al. (2008). Impact of Chess Training on Mathematics Performance and Concentration Ability of Children with Learning Disabilities. International Journal of Special Education, 23(3), 138-148. https://eric.ed.gov/?id=EJ833690 https://saintlouischessclub.org/wp-content/uploads/2025/02/Scholz_2008_Impact-1.pdf
- Ferguson, R.C. (1995). Chess in Education Research Summary. American Chess Foundation. Reference cited in: https://pmc.ncbi.nlm.nih.gov/articles/PMC4126200/
- Gaudreau, L. (1992). Etude comparative sur les apprentissages en mathematiques 5e annee. New Brunswick Department of Education.
- Liptrap, J.M. (1997). Chess and Standard Test Scores. Chess Life. Cited within Ferguson (1995) and Sala and Gobet (2016).
- Trinchero, R. (2013). Can Chess Training Improve Pisa Scores in Mathematics? An Experiment in Italian Primary Schools. Kasparov Chess Foundation Europe. http://chessedu.org/wp-content/uploads/chess_improve_pisa.pdf
- Ye, Y. (2025). Research on the Application of Chess Teaching in the Intellectual Development of Young Children. Frontiers in Psychology. DOI: https://doi.org/10.3389/fpsyg.2025.1592247
- Yakushina, N., Chichinina, E., and Dolgikh, A. (2025). Chess Classes and Executive Function in 5-6-Year-Olds. Frontiers in Psychology. DOI: https://doi.org/10.3389/fpsyg.2025.1564963
- Gevorkyan, A. et al. (2023). Chess and Mathematical Problem-Solving in Armenian Primary Schools. ASPU Scientific Research Institute of Chess.
- Rosholm, M., Mikkelsen, M.B., and Gumede, K. (2017). Your Move: The Effect of Chess on Mathematics Test Scores. PLOS ONE, 12(5), e0177257. Full text (PLOS ONE): https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0177257 PubMed: https://pubmed.ncbi.nlm.nih.gov/28494023/
- Sala, G. and Gobet, F. (2016). Do the Benefits of Chess Instruction Transfer to Academic and Cognitive Skills? A Meta-Analysis. Educational Research Review, 18, 46-57. DOI (ScienceDirect): https://doi.org/10.1016/j.edurev.2016.02.002 PDF (ResearchGate): https://www.researchgate.net/publication/296632861
- Sala, G. and Gobet, F. (2017). Does Chess Instruction Improve Mathematical Problem-Solving Ability? Two Experimental Studies with an Active Control Group. Learning and Behavior. Full text (Springer): https://link.springer.com/article/10.3758/s13420-017-0280-3
- Jerrim, J. et al. (2018). Chess in Schools: Evaluation Report. Education Endowment Foundation. Full report (ERIC): https://eric.ed.gov/?id=ED581100 PDF direct: https://files.eric.ed.gov/fulltext/ED581100.pdf
- Gao, X. et al. (2019). Visuospatial Perspective-Taking in Children Who Play Chess. Frontiers in Psychology. Search via: https://www.frontiersin.org/journals/psychology
- Ortiz-Pulido, R. et al. (2019). Neural Correlates of Chess Playing. Revista Mexicana de Neurociencia. No direct public URL verified. Cited in Gemini research data.
- Estrada-Plana, V. et al. (2024). Chess and Executive Function Development. Neuropsychologia. No direct public URL verified. Cited in Gemini research data.
- Aciego, R., Garcia, L., and Betancort, M. (2012). The Benefits of Chess for the Intellectual and Social-Emotional Enrichment in Schoolchildren. The Spanish Journal of Psychology, 15(2), 551-559. Referenced in: https://pmc.ncbi.nlm.nih.gov/articles/PMC5322219/
- Menon, V. (2016). Working Memory in Children’s Math Learning. Developmental Cognitive Neuroscience. Search via: https://www.sciencedirect.com/journal/developmental-cognitive-neuroscience
- Cubel, M. and Sanchez-Pages, S. Gender Differences in Performance Under Competition. Economic Journal. Cited in Gemini research data. No direct public URL verified.
- ISET Policy Institute. Chess in Armenian Schools: Tracking Outcomes. Institute website: https://iset.ge
- FIDE-ECU Chess in Schools Survey Report (2021). FIDE Education page: https://www.fide.com/fide/commissions/educational-chess
- Frank, A. (1970s, reviewed 1995). Chess and Cognitive Development in Zaire. Cited in Ferguson (1995).
This report was compiled by the Kingdom of Chess research team. All data gaps are explicitly flagged. Last updated April 2026.
