Data were obtained from homicide files kept by the Canadian Centre for Justice Statistics, files based largely on a special homicide form filled out by someone in the police department involved in the homicide investigation. Analysis revealed that the homicide rate among aboriginal people on reserves was roughly the same as the rate off reserves. On most dimensions examined, on- and off reserve homicides were quite similar to each other and different from nonaboriginal homicides. Homicides appeared to be intraracial; both aboriginal and nonaboriginal people tended to be killed by members of their own cultural groups. Compared to nonaboriginal homicides, aboriginal victims were more likely to be male and more likely to be killed in one- victim incidents. Aboriginal suspects, but not victims, tended to be younger and more likely to be female than nonaboriginal suspects. Male aboriginal victims, both on and off reserves, were more likely than male aboriginal victims to have been killed by a family member. Compared to homicides involving nonaboriginal victims, homicides involving aboriginal victims were more likely to have occurred after a nonviolent social encounter, a verbal argument, or a fight that escalated. In addition, consistent with other research, alcohol use by both victim and offender was more likely in aboriginal than in nonaboriginal homicides. Results suggest that the etiology of aboriginal homicides is likely to be different than that of nonaboriginal homicides, and several different explanatory concepts are discussed. 38 references, 8 notes, and 14 tables