Female sport participation is a prioritized action area in the 2018 Canadian federal budget for improving health and well-being. This study examined team sport participation prevalence and longitudinal associations with health-related behaviours among Canadian adolescent girls. We analyzed data from the COMPASS study. Participants included 1978 femalesecondary-school students who self-reported the following information at all measurement time-points (grades 9, 10, 11, and 12): socio-demographic, team sport participation status (consistent, intermittent and non-participator), moderate-to-vigorous physical activity (MVPA), dietary behaviours, smoking, cannabis use and binge drinking. Linear mixed models and generalized estimating equation models were used to examine team sport participation status in association with changes in health-related behaviours across grades. The prevalence of team sport participation declined by an average of 38.4% between grades 9 and 12. Proportions of participants being categorized as consistent, intermittent and non-participators were 25.7%, 36.4% and 37.9%, respectively. Compared to non-participators, consistent participators reported significantly greater decline in MVPA (β = - 2.77, 95% CI - 5.36, - 0.18), and the increase in odds of becoming more frequent cannabis users (OR 1.11, 95% CI 1.00, 1.26) and binge drinkers (OR 1.13, 95% CI 1.04, 1.23) was significantly greater. Girls' team sport participation had no significant longitudinal association with dietary behaviours and smoking status. Team sport participation outside of school settings is risky for prospective health-related behaviours in adolescent girls. Our findings highlight the need for investing in tailored participation initiatives that also consider how to prevent harmful substance use.