The study of child abuse and child homicide has been based on the often implicit assumption that there is a continuum of violence ranging from mild physical punishment to severe abuse and homicide. Empirical data supporting this assumption are sparse. Existing data can be shown, however, to support an assumption that there are distinct forms of violence, not a continuum. This paper reviews these data and discusses their implications for the study of violence, abuse, and homicide in terms of substantive and methodological explanations. In addition, the implications of the assumption that violence consists of distinct behaviors as opposed to a continuum are discussed in light of sociobiological and evolutionary explanations of child abuse and child homicide.