This article examines whether acculturation and experiences of discrimination help to explain observed ethnic disparities in rates of three health-compromising behaviors: interpersonal violence, drinking, and cannabis use. Data were drawn from a cross- sectional survey of 3,400 high school students from Toronto, Canada, sampled in 1998-2000. Multivariate ordinary least squares and logistic regression models tested for baseline differ- ences in the health-compromising behaviors by ethnic identity. Subsequent models adjusted for control measures and introduced acculturation and discrimination measures. Results confirm that experiences of discrimination and acculturation are risk enhanc- ing, whereas active cultural retention appears to protect ethnic youth from participation in health-compromising activities.