This research extends a 1990 study by Land, McCall, and Cohen on the structural covariates of homicide rates. Examining neighborhoods in St. Louis, this study assesses whether socioeconomic and demographic characteristics are correlated with different types of homicide, thereby addressing the question of whether homicides are sufficiently distinct in nature that their levels are not equally associated with community characteristics. The findings indicate that while residential instability is associated only with felony killings, economic disadvantage is associated with all of the homicide categories. The theoretical significance of the findings for theories of violent crime is discussed.