A contextual framework is used to examine the relationship between neighborhood characteristics and victimization while controlling for individual-level characteristics known to be associated with victimization risk. Overall, a positive relationship is found between rates of victimization and poverty, inequality, percent black, residential mobility, structural density, and percent female-headed families, independent of age, race, sex income, and marital status of victims. This pattern is stronger for theft crimes than violent crimes. Analysis of variance models were used to examine the relative predictive power of neighborhood characteristics. The results indicate that percent female-headed families, residential mobility, and structural density have the strongest effects on victimization rates. A major finding is that poverty, income inequality, and racial compositions have either weak or insignificant direct effects on victimization when percent of female-headed families is considered. In the final portion of the study, two theorems derived from Blau's (1977) macrosociological theory of social structure are tested with data on intergroup criminal victimization. The following hypotheses were tested: (1) the relative size of the same group in different neighborhood contexts is inversely related to extent of outgroup victimization; and (2) neighborhood heterogeneity is directly related to rates of intergroup victimization. The findings support these hypotheses and show that while ingroup victimization is highly prevalent in the United States, interracial and interage criminal encounters are strongly positively related to neighborhood heterogeneity. Details of the study methodology are appended, and tabular data are provided, along with about 210 references. (Author abstract modified)