["American Journal of Human Biology, Volume 38, Issue 6, June 2026. ", "\nABSTRACT\n\nObjectives\nTo characterize seasonal body composition changes in Chinese schoolchildren, contrasting in‐term periods (school‐based physical activity [PA] active) with vacations (PA suspended), and to test whether BMI, body fat percentage (BFP), and weight respond uniformly across sex and season.\n\n\nMethods\nNatural‐experiment observational analysis of a school‐monitored cohort enriched for elevated baseline adiposity (92.0% overweight or obese, Chinese WS/T 586); the research team was not involved in the school‐delivered PA program's design or delivery. We measured 342 children (212 boys, 130 girls; mean age 10.2 y) from two Beijing primary schools at up to 10 time points over 40 months (September 2021–December 2024) by bioelectrical impedance (1178 observations). Three 12‐week windows were defined (in‐term, winter vacation, summer vacation); paired changes were tested by Wilcoxon tests and era‐stratified trajectories by linear mixed‐effects models, with GAMM smooths for visualization only.\n\n\nResults\nAcross 268 in‐term paired observations, BMI declined 0.42 ± 1.10 kg/m2 (Cohen d = −0.38), an active reversal while the PA program ran. Both vacations produced gains (winter ΔBMI +0.78, summer +0.95 kg/m2). Effect sizes diverged across indicators: summer d(weight) = +1.48 > d(BMI) = +0.71 > d(BFP) = +0.57. Summer BFP rose 2.6‐fold more in girls than boys (+2.32 vs. +0.91 pp., p = 0.032).\n\n\nConclusions\nUnder this in‐term/vacation alternation, children's body composition showed marked seasonal directionality, with active reversal in term and gains during vacations. The indicator‐specific effect‐size divergence and the 2.6‐fold female‐skewed summer BFP response are consistent with sex‐stratified, tissue‐specific dynamics aligning with peri‐pubertal sexual dimorphism.\n\n"]