This study investigated attacks against transit drivers as a function of ambient temperatures from January 2006 to December 31st, 2007 in Metro Vancouver, Canada. Nonaggressive violence was differentiated from aggressive violence. Univariate, correlational analysis and ordinary least squares regression were used to determine if temperature was positively associated with attacks of violent aggression, but not nonaggressive violence. Results revealed a nonsignificant seasonal effect for violent aggression and a positive and substantial relationship between mean monthly temperature and violent aggression. No significant relationship was observed at the weekly level, possibly due to methodological limitations. The results provided some support for the heat aggression hypothesis and are interpreted using Boyanowsky’s Ecs-TC syndrome as a theoretical guide. Results also revealed support for differentiating incidents of nonaggressive violence from aggressive violence. The potential implications and future directions for research on temperature and aggression in light of climate change are discussed.