The test included direct behavioral and ecological measures of concepts central to the theory. Ecological concepts were measured at several different levels of aggregation. Of particular importance was the introduction of a control for the dangerousness of the block in which the housing unit was located. Other ecological variables included: (1) the environmental design of the housing unit (location, protective practices, single family versus other), (2) the degree of social disorganization in the neighborhood, and (3) the perceived dangerousness of the neighborhood. Measures of crucial behavioral concepts included: (1) time spent in the house during the day and (2) time spent in the house during the evening. Results revealed that none of the environmental design variables had a significant effect on victimization. The significance of the other ecological and behavioral measures differed by the type of crime. Findings have implications for refining opportunity theories of victimization. Further research is recommended. Tables, notes, and 39 references (Author abstract modified)