There is evidence of considerable variability in the quality of residential services and a tendency for residents' quality of life to co-vary with ability level. Quality of life of 154 people living in 47 small community residences was assessed using 14 subjective and objective indicators. Information on setting structure, processes, and staff working methods was collected. Factors associated with resident choice, activity, and social and community well-being were modeled using multivariate regression, which controlled for the confounding effect of ability level. Few factors associated with subjective quality of life were found. Resident ability and staff attention to residents had a strong effect across objective measures. Number of residents and staff had little explanatory power. Staff working methods had an inconsistent impact.