This study tests 3 competing hypotheses for explaining lethal violence that is conceptualized as the combination of suicide & homicide. The hypotheses differ on how lethal violence can be expressed as suicide or homicide. Suicide & homicide data from 1989-1991 were obtained from the Multiple Cause of Death Study. A bivariate analysis shows there is a positive correlation between suicide & homicide rates for 458 large counties in the United States. A weighted least square regression analysis reveals the attribution hypothesis is supported by the results. Neither the socialization nor the social disorder hypotheses are supported by the analysis. Results suggest suicide & homicide can be studied under the same theoretical & empirical framework. (30 refs.)