In the initial phase of a threefold research design, a survey was mailed to a random sample of 600 jails with capacities of between 20 and 50. Data for larger jails were obtained from a concurrent study that used a similar methodology to assess jail diversion programs for the mentally disordered. Of the 1,706 jails that received the survey, 1,053 responded. In the project's second phase, followup telephone interviews were conducted with 100 of the 149 combined mail survey respondents who had assessed their mental health services as "very effective." In the third phase, site visits were conducted for 10 jails with particularly innovative programs and policies. Findings show that most jails have no policies or procedures for managing and supervising mentally disordered detainees. Approximately 84 percent of respondents reported that mental health services were received by one-tenth or fewer of their inmates. Smaller jails (capacities of up to 99) tended to provide screening and suicide prevention services but little else, and jails with capacities of more than 1,000 offered comprehensive programs with multiple service components. Case-management services that link detainees to community services upon release were seldom provided by jails of any size. Researchers identified a number of facilities that used limited available resources to implement innovative programs and policies in six core areas: screening, evaluation, and classification procedures; crisis intervention and short-term treatment practices; discharge planning mechanisms; court liaison mechanisms; diversion practices; and contracting procedures. 4 exhibits and 8 notes