Background:Gross motor skills are postulated to have a bidirectional relationship with physical activity(PA);however,no study has tested this relationship before and after a summer break.The purpose of this study was to examine the bidirectional relationships between school PA and gross motor skills in children before and after a summer break.Methods:Participants were a sample of 440 children recruited from 3 low-income schools(age=8.9±1.2 years,mean±SD).PA was assessed as average school-day step counts using Yamax DigiWalker pedometers(Yamasa Tokei Keiki,Tokyo,Japan)worn for 5 consecutive school days.Gross motor skills were assessed using the Test for Gross Motor Development,3rd edition.Data were collected at 2 timepoints:at the end of spring semester(T1)and at the beginning of the subsequent fall semester(T2).An age-and body mass index-adjusted cross-lagged model was employed to relate T1 school step counts with T2 gross motor skills and T1 gross motor skills with T2 school step counts.Results:T1 gross motor skills significantly predicted T2 school step counts(β=0.24,95％confidence interval(95％CI):0.08-0.40,p=0.003);however,T1 school step counts did not predict T2 gross motor skills(β=0.04,95％CI:-0.06 to 0.14,p=0.445).The model explained 35.4％and 15.9％of the var-iances of T2 gross motor skills and T2 school step counts,respectively.Additional analyses indicated that these relationships were driven primarily by ball skills.Conclusion:The relationship between gross motor skills and school PA was not bidirectional;however,higher gross motor skills,specifically ball skills,predicted higher school PA after a 3-month summer break.