The cities studied were Las Vegas, Nev.; Edmonton, Alberta (Canada); Philadelphia, Pa.; Santa Barbara, Calif.; Savannah, Ga.; and Newport News, Va. The case studies showed no evidence of a single pervasive model of community policing. The six cities and their police departments varied significantly in their approach to community policing, a variance most apparent in the terminology used in each agency. These variations in forms of community policing stem from the distinctive purposes for which community policing was adopted in the different cities. The variation in community policing suggests that the concept of community policing makes sense only within the local context. Thus, the only appropriate method for evaluating community policing is to examine the change over time in a single agency with reference to achievement of its particular goals for community policing. The difficulty with the contemporary rhetoric of community policing is that clear and measurable objectives are not always specified. If objectives can be articulated, then appropriate evaluation measures can be developed to reflect the agency's progress toward meeting those objectives. Findings from the individual sites are presented in this report.