Public officials uniformly praise community policing, but the tone of their comments is bound to create high and unrealistic expectations. Although scores of police agencies in the United States and elsewhere claim to have implemented community policing, little data exists indicating that it is effectiveness. Disagreement over what represents effectiveness contributes to the current confusion. What generally passes for evaluation studies indicating success are one-group pretest-posttest designs using inadequate archival data, or poorly designed causal-comparative work. Two useful possible research designs are pretest-posttest control group design with matching, using a city of at least 60 distinct, ecologically valid neighborhoods. A second approach would be the interrupted time-series design, which entails taking measures from a single unit on the dependent variables at several points in time, introducing the treatment variable, and then measuring changes in the dependent variable at several later points. For most jurisdictions already claiming the implementation of community policing, a careful and thoughtful ex-post-facto design may be the best that can be expected. The most to hope for may be analyses that simply inform administrators, police officials, and academicians of any associations between the new policing paradigm and quality-of-life issues such as fear of crime and victimization. Appended offense definitions and 32 references