Background: Insufficient physical activity among urban youth increases risk of chronic disease.Purpose:This study assessed reported physical activity to determine when disparities in participation emerge and what ecologically guided factors are linked with high activity. Methods:We administered a cross-sectional survey to a diverse sample of 321 fourth-, sixth-, and eighth-graders measuring physical activity and a range of ecologically guided variables via self-report. Students who reported high versus low levels of physical activity were compared in bivariate and stepwise regression models stratified by gender and grade to elucidate factors associated with high physical activity. Results: The proportion of highly active students declined between fourth and sixth/eighth grade, driven largely by a decline in highly active girls. Unique factors were associated with high physical activity by grade or gender including self-efficacy among fourth-graders (adjusted odds ratio [aOR] ¼ 14.51, 95% confidence interval [CI], 1.60-131.84), social support among sixth- and eighth-graders (aOR ¼ 5.93, 95% CI, 2.04-17.29) and girls (aOR ¼ 9.03, 95% CI, 1.95-41.75), and perceived normal weight status among boys (aOR ¼ 6.62, 95% CI, 2.18-20.16). Discussion: Declines in physical activity among girls may initiate earlier than previously reported. Translation to Health Education Practice: School-based efforts to increase physical activity levels should be initiated as soon as fourth grade.