The research used an experimental design, in which police presence was increased during a 6-month period in one high-income and one low-income residential neighborhood. In addition, two high-income and two low-income residential neighborhoods were used as control groups. Information was gathered on residents' responses to both real and hypothetical cases of varied types of victimization. Results showed that the place of occurrence (public or private) and the respondent's socioeconomic status were associated with numbers of calls for police intervention. However, variables such as the perceived seriousness of the victimization or prior victimization did not show an association. In addition, calls to the police for cases of real victimization were unrelated to positive images of police effectiveness. However, a positive image of police behavior and high levels of perceived intervention by both police and neighbors were associated with a feeling of safety. Finally, discrepancies were found between real and hypothetical responses to cases of victimization. Tables and 10 references (Author abstract modified)