Early studies of black homicide were essentially descriptive and did not use multivariate analytic techniques. Additionally, criminological investigators have been less likely than other researchers to adopt modern techniques of data analysis. Most quantitative research in criminology has been conducted within two traditions: the use of spatially aggregated data for countries or regions within a country, and the study of personal attributes of criminals. Research within both these traditions may be seriously flawed, either because the data were of poor quality or because only meager mathematical tools were available to researchers. Other limitations of research on black homicide are indicative of selective inattention -- for example, inattention to the patterning of black homicide, to etiological factors, and to the situational correlates of homicide. Thus, the inadequacy of basic research on homicide and the paucity of studies on black homicide will severely impede efforts to reduce black homicide. More detailed analyses of the sociodemographic characteristics of homicide victims and offenders are needed, as are improved studies of the individual level and situational correlates of homicide. Finally, researchers must define what is meant by intervention and prevention and must determine what types of interventions work most effectively with what types of homicide. One table and 53 references are included.