Data on 9,006 households in 57 residential neighborhoods are used to estimate a model in which household burglary is a function of both attributes of individual households and characteristics of communities in which individual households are located. Research on the correlates of criminal victimization has evolved along two distinct paths. At the individual level, studies have shown that the risk of burglary varies with attributes of individual households. At the aggregate level, evidence indicates that burglary rates vary with characteristics of social areas. Results of this analysis shows that household victimization varies significantly with both characteristics of individual households and attributes of social areas. Moreover, these results provide a basis for assessing components of individual aggregate theories of victimization risk. Since household victimization varies significantly both within and between aggregate units such as neighborhoods. Models that include both household and neighborhood characteristics as independent variables are important tools for conceptually clarifying individual and aggregate theories of victimization risk. 4 tables, 9 notes, and 48 references (Author abstract modified)