Behaviour change theories have extensively been used in health behaviour change interventions and their programme theories. However, they are rarely evaluated in randomized field studies. The Let's Move It intervention targeted various psychosocial constructs to increase adolescents' physical activity. A theory-based process evaluation aiming to illuminate the trial findings as well as to test the programme theory used is conducted. Specifically, we investigate whether the intervention influenced the theorized determinants of change immediately post-intervention and after 1 year, and whether these determinants were associated with changes in physical activity. A cluster-randomized controlled trial (n = 1166). We measured theorized determinants with self-report, and physical activity (PA) with accelerometry and self-report. The effects are evaluated with repeated measures ANOVA and regression models. No changes were detected in most theorized determinants but intervention arm reported higher enactment of behaviour change techniques used during intervention immediately post-intervention and lower descriptive norms for PA throughout. Autonomous motivation was associated with PA immediately post-intervention. The lack of intervention effects may be due to many factors, for example insensitive measures, ceiling effects. However, reporting these null effects advances understanding of behaviour change processes. We introduce methodologic possibilities for future intervention programme theory evaluation efforts.