Abstract Background This study aimed to evaluate the effectiveness of obesity prevention educational activities, isolated or combined with the modification of the school environment on total physical activity time. Methods This is a school-based randomized controlled trial, conducted with 2511 students from fifth and sixth-grade in Brazil, that employed a parallel, three-group experimental arms: control group (CG), PAAPPAS group (PG) and PAAPPAS-environment group (PEG). During the 2016 school year, the PG received educational activities in the classroom, providing a general basis for a healthy lifestyle. The PEG received the same educational activities and also modifications in the school environment to stimulate physical activity practice during one month. Physical activity was evaluated through a validated physical activity questionnaire. Intention-to-treat analysis was performed to evaluate the rate of change of physical activity time between groups using PROC GENMOD procedure. All analyses were performed using SAS. Results Total physical activity time increased by 22% in PEG compared with CG (P = 0.003). No difference was observed between PEG and PG (Δ = 60.20 vs. 36.37, respectively; P = 0.27) and between PG and CG (Δ = 36.37 vs. 9.70, respectively; P = 0.23). The proportion of individuals who attended at least 150 min week−1 of physical activity increased in PEG compared with PG (P = 0.04); however, no difference was observed between PEG and CG (P = 0.19) and between PG and CG (P = 0.26). For 300 min week−1, no difference was observed between groups. Conclusion A school-based multi-component intervention including modification of the school environment was effective for increasing physical activity time among adolescents.