Data were collected in early 1996 from all 53 distinct neighborhoods within the city. Information about organized neighborhood watch groups and community-based organizations was obtained by means of telephone interviews with community police officers and community-oriented policing sergeants in all police districts. Community characteristics that are known or hypothesized determinants of crime were used as control variables. The dependent variable was the total number of reported crimes in each community. Data were analyzed using ordinary least squares regression techniques. Results confirmed previous studies demonstrating that increases in poverty, mobility, racial heterogeneity, and population density are associated with increases in crime in a community. However, the existence of neighborhood watch groups that interact with the police was not a good predictor of total crime. In addition, an increased number of community-based organizations was associated with an increased level of crime in the community. Findings support prior research findings that question the effectiveness of collective crime prevention efforts. Findings suggested that neighborhood watch or individual crime prevention efforts have no effect on crime. The finding that larger numbers of community-based organizations were associated with higher amounts of crime suggested the need for further research to clarify the reasons. Table, notes, and 39 references (Author abstract modified)