This Note considers future directions for health care quality assessment research and the uses to which its products should be put. Health services research can profitably be focused on issues of efficacy, effectiveness, variations in use, or quality of care. The authors argue, however, that future important issues require that information from all these areas be integrated into a "macro" model that will address continuing problems in the medical system. They identify a number of research needs in the quality assessment field: (1) establish the links between the process and outcomes of care, (2) establish the clinical validity of process measures, (3) develop outcome measures that are feasible to collect, (4) continually refine definitions of patient outcomes based on evolving clinical understandings of disease, (5) learn more about physician-patient communication and interaction, (6) develop an epidemiology of the quality of medical care, and (7) measure the direct and indirect costs and benefits of medical care. The authors conclude that we need a fully integrated system of data collection and information dissemination and we must discard medical procedures and services that are not efficacious.