Context Ice hockey has one of the highest sport participation and injury rates in youth in Canada. Body checking (BC) is the predominant mechanism of injury in leagues where it is permitted.

Objective To determine if the incidence of physical contacts differs for Pee Wee (PW) players (11–12 years) in leagues where BC is permitted (Alberta) versus players in leagues where BC is not permitted (Quebec).

Design Cohort study.

Setting Arenas in Alberta and Quebec, 2007–2008 season.

Participants PW teams in the top 10% of divisions of play in Quebec (n=10) and Calgary (n=10).

Observation technique 20 games were analysed. Games were video-recorded and analysed by two trained observers.

Main outcome measures The intensity of the physical contacts was observed with a validated observation system. Five levels of intensity were coded. Level 1 represents the lowest intensity of physical contacts. Rate ratios (RR) adjusted for team clusters based on multivariate Poisson regression was used to compare games played in Calgary and Quebec.

Results 2418 physical contacts were observed. No difference was observed in the rate of physical contacts in games played in both provinces. High intensity physical contacts (levels 4 and 5 combined) were more frequent in Calgary (RR=6.57; 95% IC 3.09 to 13.93).

Conclusions The results suggest that the greater rate of injury observed in PW leagues where BC is permitted would not be caused by the frequency of the physical contacts but by their intensity.