Homeless people experience significant, persistent inequities in health and access to health care services compared to the general population. When a homeless person has a concurrent mental health and substance use disorder (concurrent disorder), it puts them at even more disadvantage. The prevalence of concurrent disorders in the homeless population is estimated to be 10-20 percent across North America, although the actual prevalence may be far greater (BC Ministry of Social Development and Economic Security, 2001). In a 2007 community consultation, inner city health service providers described the need for better, more integrated and more responsive services for their homeless clients with concurrent disorders. To help guide policy and program planning, providers said a stronger evidence base was needed to explain what types of community-based interventions best retain and engage clients, and reduce their problems with mental health problems and substance use. To respond to this request, our interdisciplinary research team carried out a realist-informed, community-partnered evidence synthesis in 2007-2008. We took an integrated knowledge translation approach to promote research uptake and implementation. Community partners were involved in all aspects of the project, including identifying the research question, determining the scope of review, assessing the relevance of literature, and disseminating results. Project partners were the Centre for Research on Inner City Health; Access Alliance Multicultural Health and Community Services; the Ontario Federation of Indian Friendship Centres; Sistering, a Women's Place; Street Health; and South Riverdale Community Health Centre in Toronto. Our goal was to advise the Ontario Ministry of Health and Long-Term Care and Local Health Integration Networks about appropriate services for homeless people with concurrent disorders. We also wanted to show how grassroots collaborations can contribute to positive health and social service system change, and identify a new method for synthesizing population health intervention research evidence.