We assessed which intervention components were associated with change in moderate-to-vigorous physical activity (MVPA) and wellbeing through proposed psychosocial mediators. Eight schools (n = 1319; 13−14 years) ran GoActive, where older mentors and in-class-peer-leaders encouraged classes to conduct two new activities/week; students gained points and rewards for activity. We assessed exposures: participant-perceived engagement with components (post-intervention): older mentorship, peer leadership, class sessions, competition, rewards, points entered online; potential mediators (change from baseline): social support, self-efficacy, group cohesion, friendship quality, self-esteem; and outcomes (change from baseline): accelerometer-assessed MVPA (min/day), wellbeing (Warwick-Edinburgh). Mediation was assessed using linear regression models stratified by gender (adjusted for age, ethnicity, language, school, BMI z-score, baseline values), assessing associations between (1) exposures and mediators, (2) exposures and outcomes (without mediators) and (3) exposure and mediator with outcome using bootstrap resampling. No evidence was found to support the use of these components to increase physical activity. Among boys, higher perceived teacher and mentor support were associated with improved wellbeing via various mediators. Among girls, higher perceived mentor support and perception of competition and rewards were positively associated with wellbeing via self-efficacy, self-esteem and social support. If implemented well, mentorship could increase wellbeing among adolescents. Teacher support and class-based activity sessions may be important for boys’ wellbeing, whereas rewards and competition warrant consideration among girls.