The main purpose of the proposed research is to continue evaluation of the effects of a McKinney Research Demonstration Program for the Homeless. The San Diego Project was one of six McKinney program sites funded by NIMH in October, 1990. Its purpose was to evaluate the effectiveness of providing comprehensive supportive services coordinated with independent housing alternatives for severely and persistently mentally ill homeless persons. A total of 362 clients were recruited to the project and were randomly assigned to four conditions: (l) comprehensive case management (CCM) and housing; (2) traditional case management (TRDCM) and housing; (3) CCM and no housing; and (4) TRDCM and no housing. By the end of year 03 (9/93), the project will have collected baseline and six and twelve month followup interview data on all subjects and eighteen month followup data on approximately 50 percent. Monthly case manager reports and case file abstraction data will have been collected on all respondents for at least twelve months. This proposal is to continue evaluation followup through three years post- project entry for all subjects. The primary research hypothesis is the same as in the original grant - that the combination of comprehensive and flexible supportive services with permanent housing will produce better client outcomes including: a) level of severity of psychopathology; b) increased stability of housing; c) improved functional status; d) improved quality of life and life satisfaction; e) improved physical health condition; and f) more effective use of mental health services. More specifically, researchers hypothesize that comprehensive support services and housing will each produce better outcomes than the comparison condition of traditional homeless case management. The design allows testing for the additive effects of the two experimental conditions and for interactive effects as well. The proposed project also intends to evaluate the effects of a """"""""step down"""""""" of service level for those originally assigned to the CCM conditions (l and 3). Little is known concerning whether original gains achieved in CCM are maintained when the level of support is decreased or removed. When clients in those two conditions finish eighteen months of comprehensive case management, they are being randomly assigned to (a) a continuation of the original comprehensive case management for another 12 months or (b) a transitional case management condition which closely resembles traditional homeless case management. The same outcomes as above will be tracked.