These innovations should indicate whether the social or institutional context of specific activities affects the risk of victimization independent of the situational activity. Moreover, they should facilitate determination of the causal link between activity and victimization. The survey used to collect the study data was the Victim Risk Supplement to the National Crime Survey. This supplement was organized in activity domains: at work, at school, at home, and at leisure. Both routine activities and victimizations were classified in these domains. One of the surprising results of the analysis was the large proportion of crimes that occurred at work. Only approximately 50 percent of all crime occurred at or near the victim's residence. Victimization rates for school-aged persons at school were three times those of school-aged persons in other activity spheres. The patterns of risk rates across race and income groups also suggest that the amount of discretion available in an activity domain many affect risk therein. 2 tables and 21 references