In an effort to address the methodological issues associated with units of analysis, the authors offer a systematic review that examines the effects of structural factors on aggregated, disaggregated, and time-series homicide studies. Of central concern is the comparison of theoretically relevant structural variables used in homicide studies across various levels of analysis. The chapter begins with a summary of criminological approaches and the identification of those theoretical variables that the authors will compare across studies. The focus is on three theoretical approaches most commonly used in the study of homicide: social disorganization theory, strain/deprivation theory, and the southern subculture of violence thesis. Next, the chapter summarizes the salient empirical research that has examined covariates of homicide rates during various periods and across various levels of aggregation. Inconsistent findings are explained across these studies by examining statistical characteristics of the data that may account for these discrepancies. The chapter concludes with recommendations that may help researchers avoid the production of erroneous substantive inferences in future research. 3 tables, 4 notes, 58 references, and appended supplementary information on the studies reviewed