IN THE FIRST PHASE, A NATIONAL SURVEY OF 118 JAILS PROVIDED DATA ON SCREENING, DETOXIFICATION, EVALUATION, TREATMENT, AND REFERRAL SERVICES FOR DRUG-ADDICTED JAIL INMATES. ON THE BASIS OF THESE DATA FOUR ORGANIZATIONAL MODELS OF SERVICE DELIVERY WERE DEVELOPED: (1) INTERNAL (ALL SERVICES ADMINISTERED AND DELIVERED BY JAIL PERSONNEL) (2) INTERSECTION (SERVICES PROVIDED BY COMMUNITY AGENCIES IN COOPERATION WITH JAILS), (3) LINKAGE (EMPHASIS ON REFERRING INMATES TO COMMUNITY AGENCIES UPON RELEASE FROM JAIL), AND (4) COMBINATION (MAY INCLUDE BOTH INTERNAL AND EXTERNAL SERVICE DELIVERY, AND EITHER LINKAGE OR INTERSECTION). IN THE SECOND PHASE, THE FOUR MODELS WERE VALIDATED IN FIVE URBAN JAILS (TWO REPRESENTATIVE OF THE COMBINATION MODEL). THE JAILS WERE IN DETROIT, MICH., PHOENIX, ARIZ., BIRMINGHAM, ALA., BUFFALO, N.Y., AND SAN DIEGO, CALIF. A CASE STUDY IS PRESENTED FOR EACH JAIL AND ORGANIZATIONAL MODEL, WITH EMPHASIS ON HOW THE MODEL WORKS; ADVANTAGES AND DISADVANTAGES FOR JAIL ADMINISTRATION, COMMUNITY AGENCIES, AND INMATES; AND LIMITATIONS COMPARED TO OTHER DELIVERY SYSTEMS. THE SECOND PHASE REPORT ALSO EXAMINES POLICY ISSUES PERTAINING TO HOW STATE AGENCIES CAN INFLUENCE THE PROVISION OF DRUG ABUSE SERVICES IN LOCAL JAILS. POLICY RECOMMENDATIONS ARE PRESENTED, TOGETHER WITH GENERAL GUIDELINES FOR PROVIDING DRUG TREATMENT TO JAIL INMATES IN ANY ORGANIZATIONAL FORMAT (CLEARLY DEFINED GOALS, TRAINING FOR JAIL STAFF, COMPREHENSIVE SCREENING, GOOD COMMUNICATION AND COORDINATION, ETC.). A SCHEMATIC DIAGRAM OF EACH MODEL IS PROVIDED. (LKM)