This study tests components of Wong and Solomon's (2002, Mental Health Services Research, 4(2), 13-28) model of community integration, identifying both the dimensions and predictors of integration. It evaluates community integration among adults with psychiatric disabilities assigned randomly to receive either independent scatter-site apartments with the Housing First approach (experimental) or services as usual (control). Factor analysis supported a definition of community integration that includes psychological, physical, and social domains, but also suggested the existence of another factor, independence/self-actualization. Regression analysis suggested that choice and independent scatter-site housing were predictors of psychological and social integration respectively. Psychiatric hospitalization, symptomatology and participation in substance use treatment were also found to influence aspects of integration. We discuss several issues that future studies should explore including the possibility that the same factor can differentially influence discrete aspects of integration, the role of person-environment fit, integration that is not based in the neighborhood, and, finally, conceptions of community integration from the perspective of consumers themselves. (abstract from http://www.ingentaconnect.com)