BackgroundA gender-based disparity in physical activity (PA) among youth, whereby girls are less active than boys is a persistent finding in the literature. A greater understanding of the mechanisms underlying this difference has potential to guide PA intervention strategies.MethodsData were collected at age 8 and 12 years (276 boys, 279 girls) from 29 schools as part of the LOOK study. Multilevel linear models were fitted separately for boys and girls to examine effects of individual, family and environmental level correlates on pedometer measured PA. Cardio-respiratory fitness (multi-stage run), percent fat (DEXA), eye-hand coordination (throw and catch test) and perceived competence in physical education (questionnaire) were used as individual level correlates. At the family level, parent’s support and education (questionnaire) were used. School attended and extracurricular sport participation were included as environmental level correlates.ResultsGirls were 19% less active than boys (9420 vs 11360 steps/day, p<0.001, 95%CI [1844, 2626]). Lower PA among girls was associated with weaker influences at the school and family levels and through lower participation in extracurricular sport. School attended explained some of the variation in boys PA (8.4%) but not girls. Girls compared to boys had less favourable individual attributes associated with PA at age 8 years, including 18% lower cardio-respiratory fitness (3.5 vs 4.2, p<0.001, CI [0.5,0.9]), 44% lower eye-hand coordination (11.0 vs 17.3, p<0.001, CI [5.1,9.0]), higher percent body fat (28% vs 23%, p<0.001, CI [3.5,5.7]) and 9% lower perceived competence in physical education (7.7 vs 8.4, p<0.001, CI [0.2,0.9]). Participation in extracurricular sport at either age 8 or 12 years was protective against declines in PA over time among boys but not girls.ConclusionGirls PA was less favourably influenced by socio-ecological factors at the individual, family, school and environmental levels. These factors are potentially modifiable suggesting the gap in PA between boys and girls can be reduced. Strategies aiming to increase PA should be multicomponent and take into consideration that pathways to increasing PA are likely to differ among boys and girls.