Background <p>The dose-response effects of physical exercise on brain volume remain insufficiently understood, and it is yet to be established whether exercise-induced relative changes in brain volume translate into improvements in cognitive function. Furthermore, it remains unclear whether specific physical adaptations are associated with both exercise-related changes in brain volume and cognitive outcomes. </p> Methods <p>To address these gaps, we conducted a pre-registered systematic review and multilevel meta-analysis of 59 randomized controlled trials comprising a total of 5,099 participants. Comprehensive literature searches were carried out in PubMed, PsycINFO, Embase, and Web of Science up to August 2024. The meta-analysis examined overall effects of physical exercise on brain volume, investigated dose-response effects across relevant moderators, and explored potential mediation pathways linking exercise-induced relative changes in brain volume to cognitive outcomes.</p> Results <p>This study found that physical exercise had a significant effect on brain volume. Dose-response analyses further revealed that this effect was more pronounced among individuals of older age, with longer intervention durations and higher compliance. Exercise sessions lasting 40–60&#xa0;min were associated with the most robust effects on brain volume. Post-hoc analyses suggest that these effects are more likely attributable to the preservation of brain volume rather than substantial volumetric increases. In addition, regression analysis demonstrated a significant positive association between exercise-induced relative changes in brain volume and cognitive function. Moreover, mediation analyses indicated that exercise-induced physical adaptations in muscle function, motor performance, and cardiorespiratory fitness were indirectly associated with cognitive performance through their effects on relative brain volume change.</p> Conclusions <p>These findings indicate that physical exercise exerts beneficial effects on brain volume, particularly under conditions such as older age, sufficient exercise duration, and high compliance. Exercise-induced relative changes in brain volume were significantly associated with cognitive function and served as a mediator between physical adaptations and cognitive outcomes. Together, these results provide a mechanistic foundation and practical insights for designing targeted exercise interventions to promote brain health in aging populations.</p> Trial registration <p>PROSPERO CRD42024525635.</p>