These data were collected in Oakland, California, and Birmingham, Alabama, to examine the effectiveness of alternative drug enforcement strategies. A further objective was to compare the relative effectiveness of strategies drawn from professional- versus community-oriented models of policing. The professional model emphasizes police responsibility for crime control, whereas the community model stresses the importance of a police-citizen partnership in crime control. At each site, experimental treatments were applied to selected police beats. The Oakland Police Department implemented a high-visibility enforcement effort consisting of undercover buy-bust operations, aggressive patrols, and motor vehicle stops, while the Birmingham Police Department engaged in somewhat less visible buy-busts and sting operations. Both departments attempted a community-oriented approach involving door-to-door contacts with residents. In Oakland, four beats were studied: one beat used a special drug enforcement unit, another used a door-to-door community policing strategy, a third used a combination of these approaches, and the fourth beat served as a control group. In Birmingham, three beats were chosen: Drug enforcement was conducted by the narcotics unit in one beat, door-to-door policing, as in Oakland, was used in another beat, and a police substation was established in the third beat. To evaluate the effectiveness of these alternative strategies, data were collected from three sources. First, a panel survey was administered in two waves on a pre-test/post-test basis. The panel survey data addressed the ways in which citizens' perceptions of drug activity, crime problems, neighborhood safety, and police service were affected by the various policing strategies. Second, structured observations of police and citizen encounters were made in Oakland during the periods the treatments were in effect. Observers trained by the researchers recorded information regarding the roles and behaviors of police and citizens as well as police compliance with the experiment's procedures. And third, to assess the impact of the alternative strategies on crime rates, reported crime data were collected for time periods before and during the experimental treatment periods, both in the targeted beats and city-wide.