To evaluate the effectiveness of a behaviourally focused nutrition education (NE) intervention based on the Health Belief Model (HBM) to improve knowledge, attitudes and practices (KAP) related to eating habits and activity levels in 10–12-year-old adolescents in Mumbai, India.

School-based cluster randomised controlled trial. The experimental group (EG) received weekly NE and three parent sessions over 12 weeks; no sessions were conducted for the control group (CG). The theoretical framework of HBM and focus group discussion results guided the development of behaviour change communication strategies and NE aids. KAP were measured using a validated survey instrument, administered at baseline and endline in EG and CG. Paired and independent t tests determined within-group and between-group changes in pre–post scores.

Two aided and two private schools that were randomly allocated to either an EG or CG.

Adolescent boys and girls (n 498; EG n 292 and CG n 206).

EG reported improvements in mean knowledge (39·3%), attitude (7·3 %), diet (9·6 %) and activity practice (9·4%) scores from pre to post intervention. No significant changes were observed in CG. Significant improvements in scores associated with perceived benefits, barriers and self-efficacy, breakfast and vegetable consumption, and moderate-to-vigorous activities were observed in EG.

Integrating NE into the academic curriculum and adopting evidence-based lessons that entail targeted information delivery and participatory activities can improve knowledge, foster right attitudes and facilitate better eating and activity-related practices in Indian adolescents.