While regional differences in homicide rates have been well established, explanations for the phenomenon have been elusive. Two general interpretations of these regional differences in homicide are discussed. First, the subcultural thesis, articulated most completely by Raymond Gastil, states that predisposition to lethal violence in southern regional culture . . . accounts for the greater part of the relative heights of the American homicide rate. This theory, however, has been challenged. Critics of the subcultural theory claim that the effects of nonculturual variables had been seriously underestimated by a failure to provide adequately sensitive control measures. Consequently, they constructed a structural-poverty index, which considers the effects of poverty; this index was used in a regression equation containing other variables. The dependent variables of the study were primary homicide rates and nonprimary rates; independent variables were basically updated versions of those used in the Loftin and Hill study. Results reconfirm the importance of social-structural variables in predicting homicide rates; further, they support Loftin and Hill's refutation of subcultural explanations for regional differences in rates of homicide. Results suggest that conditions (both social and psychological) associated with poverty in the United States increase the tendency for economically disadvantaged individuals to commit homicide. However, poverty is less important in explaining variations in nonprimary homicide rates. Additional research with more control of situational factors is warranted. Tables, footnotes, and 33 references are included.