Case-control studies and other analytic methods are needed to test causal hypotheses regarding the risk of homicide victimization or perpetration for individuals. Strengths and limitations of the case-control method are illustrated by comparing the design with cohort studies. Fundamental issues include the selection and comparability of cases and controls, effects of bias, interpretation of risk estimates, and problems of implementation. The article presents the findings of a recently published study on the use of the case-control method to examine 22 unsolved child homicides and 2 unsolved disappearances of children in Atlanta, Ga., between July 1, 1979, and March 15, 1981. Increased use of this method should advance our understanding of homicide and other forms of intentional violence. Tabular data and 40 references. (Author abstract modified)