The main purpose of the present study is to demonstrate the structure, mechanisms, and efficacy of community policing and its impact on perceived disorder, crime, quality of life in the community, citizens' fear, and satisfaction with the police. It compares traditional and community policing paradigms on three dimensions: goal, measurement of outcome, and approach to crime. It concludes that community policing has a comprehensive, community-oriented goal, targets both disorder and crime, and emphasizes both organizational and community measures in police evaluation. It also addresses the criticisms of community policing and tests the heatedly debated relationships concerning community policing, disorder, crime, citizen's fear, and collective efficacy. The major findings of the study include (1) Harcourt's falsification of Skogan's findings is invalid because of the methodological flaws, and therefore does not negate the disorder-crime nexus; (2) Sampson and Raudenbush unintentionally demonstrate, through their reciprocal feedback models, that crime and disorder are indirectly related; (3) disorder has strong direct, indirect, and total effects on crime even with collective efficacy being controlled for; (4) contrary to intuition, disorder elicits more fear than crime; (5) community policing reduces crime indirectly; (6) collective efficacy plays a far less significant role in controlling disorder, crime, and fear than community policing; and (7) citizens' fear and perceived life quality are significant predictors of citizen satisfaction with the police. (Contains 3 figures, 1 table, and 11 notes.)