Provider capitation may constrain costs, but it also may reduce access and quality of care. We examine the impacts of capitating the usual source of care of enrollees in health maintenance organizations (HMOs). We account for the endogeneity of capitation and other characteristics using generalized methods of moments (GMM) estimation on a sample from the Medical Expenditure Panel Survey for 1996 and 1997. Being organized as a group/staff HMO generally has stronger impact on access and quality than capitation. Capitation by itself may increase access to consumers' usual sources of care, improve primary preventive care, and reduce coordination, but estimates with GMM were not statistically significant.