Interventions to increase physical activity in children have adopted broad approaches and achieved varying success. There is a need to adopt approaches underpinned with a theoretical basis. Accordingly, the aim here was to implement and evaluate a 12-week intervention designed using the concepts of the COM-B model to determine the effect this has on physical activity levels. One hundred and forty-seven school-age children (mean age 8.9 ± 1.3 years) took part in a 12-week program delivered in a school setting. Topics included physical activity, healthy eating, sleep quality and reducing screen time/sedentary activities when not in school. A sample of participants wore a wrist-worn accelerometer for seven days pre-and post-intervention (N = 11). The physical activity frequency was unchanged (2.9 ± 1.0 AU) when compared with post-intervention values (3.1 ± 0.8 AU, mean increase 6.8 ± 3.7%, p > 0.05). Changes were observed in the daily consumption of fruit and vegetables (pre-intervention 44.6% vs. post-intervention 60.2%, p < 0.05). Sedentary time, light activity, moderate activity and vigorous activity were unchanged post-intervention (p > 0.05). There is a need to adopt a broader approach that incorporates a theoretical basis and considers the complex ways by which physical activity behaviours are influenced.