The data came from Hayward, Calif., Davenport, Iowa, Ann Arbor, Mich., Sioux City, Iowa, Pocatello, Idaho, and Ontario, Calif. Study information came from self-administered mail surveys, direct observations, and interviews. Results revealed that the concepts of sense of community and social capital helped explain the strength of a community’s social fabric and that the relationship between levels of social capital and programmatic activity in community oriented policing is curvilinear rather than linear. Thus, community policing may act as either a change agent or a partner in coproduction of order, depending on the community context; community factors may have more influence than police factors on community policing. Findings also suggested that community policing may have two distinct evolutionary stages, one for communities with low levels of Putnam’s social capital and the other when police agencies can increase the stocks of social capital and enhance citizen participation in community policing programs. Findings also suggested that the number of participants actively engaged in a community policing program is not as important as the character of the individuals participating in the program. Tables, footnotes, appended methodological information and tables, and 149 references