Although mental health court has been found to increase quality of life and functioning, it is unknown whether community stability indicators predict program completion and delay re-arrest for homeless versus non-homeless mental health court participants. Self-reported quality of life and social support, chart diagnosis, and administrative housing, services and criminal justice data were collected for 589 Bronx Mental Health Court participants for 12 months following diversion. Individuals who were homeless (n = 89) versus housed (n = 500) during the 12 months prior to diversion were compared to determine predictors of court graduation and rearrest. Homeless status did not predict graduation or re-arrest, and quality of life, social support, functioning and housing type were not predictive of re-arrest. Housing instability negatively affected outcomes for both homeless and non-homeless individuals. An increase in functioning and a slight improvement in social support were predictive of successful graduation for homeless individuals,