The study examines boundary spanning activity between Hidden Entrance, a crisis outreach program in Prince Georges County, Md., and the county police departments. The program provides around the clock counseling in such areas as family crisis, runaways, and children in need of supervision (CINS). The program relies heavily on outside agency referrals. The treatment strategies are conceptualized in terms of temporary and permanent linkage systems. The first strategy focuses on a single mode of communication (e.g., an oral presentation) on a one-time basis. The second strategy uses boundary spanners in the development of a permanent linkage system and incorporates several transfer media used on a continuous basis. To test the treatment strategies, a randomized block design has been employed. By random assignment, two groups have been created, each consisting of one State police barracks, two local departments, one special department, and nine municipal departments. Research findings are based on data obtained from Hidden Entrance's records of police use of its services. The study shows no significant differences between treatment groups, in terms of the number of police referrals to Hidden Entrance. However, several qualitative and theoretical issues stem from the study. For example, a meaningful relationship between police and the program can be achieved through instilling more confidence among police relative to assistance provided by the program staff. To this end, the staff should consistently inform the police about outcomes of cases referred by the police. In demonstrating the use of experimental research methods, the study draws attention to the problems of implementing any 'different' strategy in an applied research setting. The importance of random assignment, in this case aimed at overcoming selection bias, becomes crucial when the experimental strategy is undertaken in the 'real world'. Statistical data, a discussion of previous research, notes, and about 30 references are included.