County homicide rates, the dependent variable, were obtained from vital statistics compiled by the National Center for Health Statistics for 1979 through 1982. Cumulated data were used to stabilized the rates of annual incidence because of year to year fluctuations. Logarithmic transformation was performed because of skewed distribution in homicides where some counties did not report a single homicide and others reported high numbers. The independent variables, with the exception of religious diversity, were based on tape data from County Statistics File 1 (United States Bureau of the Census 1984). Religious data were obtained from the Roper Center. Ordinary least squares analysis was used to assess the relative contribution of the series of independent variables to the explanation of the dependent variables. A zero-order correlation matrix shows the relationship of all independent variables to the dependent variable of the homicide rate. Three bivariate relationships were determined to exhibit multicollinearity: the percent of the population that is rural and the divorce rate; the percent of the population that is rural nonfarm and the divorce rate; and median years of education and the percent of families living below poverty. The results are in accordance with Durkeim's theory which claims that less developed areas facilitate more traditional bonds within a society and thus reduce homicide rates. 2 tables and 63 references