Objective To assess the reliability and validity of 3- to 6-year-old Children's General Physical Activity Scale, so as to provide reference for the establishment of national norm and clinical application. Methods With stratified cluster random sampling method, 210 children from 3 kindergartens in Nanjing were tested to analyze and evaluate the reliability and validity of the scale.Cronbach's α coefficient, half reliability coefficient, test-retest reliability coefficient and intraclass correlation coefficient (ICC) were used to evaluate the internal and external reliability of the scale.Exploratory factor analysis was used to evaluate the construct validity of the scale.The content validity was evaluated by expert argumentation and related analysis.Using Peabody Developmental Motor Scales-2 (PDMS-2) as the calibration standard, the correlation between the scores of the two scales in 30 children was analyzed to evaluate the correlation validity of the calibration standard. Results The Cronbach's α value of the scale was 0.815, the Guttman half-split coefficient was 0.804, the test-retest reliability was 0.91, and the rater reliability was 0.994.The KMO measure of sampling adequacy was 0.783.The correlation coefficient between total score and each dimension score ranged from 0.596 to 0.721.Except the body flexibility, the ability of children in each item showed an increasing trend with age (F=6.070—65.355,P<0.01).The total standard score of 3- to 6-year-old Children's General Physical Activity Scale was positively correlated with the standard score of PDMS-2 (r=0.552, P=0.002).There were significant differences in each dimension between the typically developed group and the motor delay group (t=4.536—16.072, P<0.01). Conclusion 3- to 6-year-old Children's General Physical Activity Scale has acceptable reliability and validity, reaching the basic psychometric demands, which can be used to make the norm.