What began as an experiment in five police districts has expanded to encompass the entire city. Under CAPS, teams of officers now have relatively long-term assignments in each of the city's 279 police beats. The entire department has been trained in problem solving under the CAPS five-step process, which focuses on victims, offenders, and crime locations. The problem solving efforts of beat officers are supported by a coordinated system for delivering city services. The commitment to community involvement is reflected in beat meetings and district advisory committees. The first section of this report summarizes what has been learned about citizen involvement in the program based on an analysis of beat meetings and district advisory committees. This is followed by a description of changes over time in city residents' assessments of the quality of police service. Trends in crime and fear in city neighborhoods are described, followed by a detailed description of CAPS problem solving efforts and trends in neighborhood problems. The report concludes with an in-depth appraisal of conditions in the Latino community. Regarding public involvement, this report concludes that beat meetings have failed to produce ongoing coordinated officer-citizen actions that impact the community. Issues raised at beat meetings are not well represented in officer reports, such that little information exists on issues raised by citizens and what has been done to address them. Partnerships among city agencies in implementing the problem solving strategy has been excellent, and the city has effectively reorganized its institutions to support community policing. The report concludes that problem solving in practice, however, has been the least effective component of CAPS. Effective problem solving to reduce crime requires extensive training, close supervision, a strong analytical capacity, and organization-wide commitment. Efforts to solve local priority problems have not been effective. The same problems in the same locations persist year after year. Training in and the monitoring of problem solving is lacking. A more systematic and intensive approach to the problem solving process could also re-engage the community in the active partnership promised by CAPS. 1 table and 7 figures