Background: School environments may promote or hinder physical activity in young people. The purpose of this research was to examine relationships between school recreational environments and adolescent physical activity. Methods: Using multilevel logistic regression, data from 7638 grade 6 to 10 students from 154 schools who participated in the 2005/06 Canadian Health Behaviour in School-Aged Children Survey were analyzed. Individual and cumulative effects of school policies, varsity and intramural athletics, presence and condition of fields, and condition of gymnasiums on students' self-reported physical activity ([greater than or equal] 2 h/wk vs less than 2 h/wk) were examined. Results: Moderate gradients in physical activity were observed according to number of recreational features and opportunities. Overall, students at schools with more recreational features and opportunities reported higher rates of class-time and free-time physical activity; this was strongest among high school students. Boys' rates of class-time physical activity were 1.53 (95% confidence interval (CI) = 1.12-1.80) times as high at high schools with the most recreational features as at schools with the fewest. Similarly, girls' rates of free-time physical activity at school were 1.62 (95% CI: 0.96-2.21) times as high at high schools with the most opportunities and facilities as compared to schools with the fewest. Modest associations were observed between individual school characteristics and class-time and free-time physical activity. Conclusions: Taken together, the cumulative effect of school recreational features may be more important than any one characteristic individually. (Contains 4 tables and 1 figure.)