Evaluated the second of the 2 core questions around which the ACCESS (Access to Community Care and Effective Services and Supports) evaluation was designed: Does better integration of service systems improve the treatment outcomes of homeless persons with severe mental illness? The ACCESS program provided technical support and about $250,000 a year for 4 yrs to 9 sites to implement strategies to promote systems integration. These sites, along with 9 comparison sites, also received funds to support outreach and assertive community treatment programs to assist the 100 clients a year at each site. Outcome data were obtained at baseline and 3 and 12 mo later from 7,055 clients across 4 annual cohorts at all sites. Clients at all sites showed improvement in outcome measures. More extensive implementation of systems integration strategies was unrelated to these outcomes. However, clients of sites that became more integrated, regardless of the degree of implementation or whether the sites were experimental sites or comparison sites, had progressively better housing outcomes. It is concluded that interventions designed to increase the level of systems integration in the ACCESS demonstration did not result in better client outcomes.