The study found that the frequency of police attendance at community meetings was negatively related to informal social control after controlling for neighborhood demographics, crime, and social cohesion. Police attendance at community meetings was most frequent in disadvantaged neighborhoods with little social cohesion and where residents had generally negative perceptions of police. Also, few residents attended community meetings, which limited any constructive impact of police-citizen cooperation. The strongest predictors of informal social control were neighborhood social cohesion and government responsiveness to neighborhood needs. These factors were also significantly related to perceptions of police effectiveness and legitimacy. All of these factors were weak in disadvantaged neighborhoods, and they were apparently not changed by efforts at community policing. More research is required to bring understanding of how negative perceptions of police and government can be addressed in the most disadvantaged communities. The study surveyed 81 neighborhood leaders who represented 81 neighborhoods in Portland, OR. The dependent variable was informal social control. It was measured by asking community leaders about the willingness of residents in their neighborhoods to intervene in specific incidents of antisocial and delinquent/criminal behavior in the neighborhood. Interviews with the community leaders also produced information on independent variables related to community policing. These variables were frequency of police attendance at community meetings, police accessibility, police-resident problem solving, police effectiveness, and residents' fear of police encounters. Census and crime data were used to measure independent variables related to structural disadvantage. Control variables were fear of crime, social cohesion, and government responsiveness. 3 tables, 4 notes, and 60 references