The health care delivery system is being restructured with the immediate goal of containing costs. Thus, the success of competing systems is typically equated with how well they achieve this goal. But in the long run, changes in the organization and financing of care for the elderly must strike a balance between containing costs and preserving the features of medical care process essential to producing good outcomes. This situation is particularly crucial for the elderly, who are being encouraged to enroll in health maintenance organizations and competitive medical plans to control rising costs under the Medicare program. To control costs without harm to health, policymakers must be better informed about the effects of cost containment of the style of medical care practiced and the implications for patient outcomes. We propose to analyze data from the Medical Outcomes Study (MOS), an ongoing investigation that provides a unique opportunity to address these issues. The MOS is a major 6-year study of patients with prevalent and treatable chronic conditions (hypertension, diabetes, heart disease and depression) sampled to represent solo and group practices and both prepaid and fee-for- service arrangements in Boston, Chicago, and Los Angeles. The MOS database is characterized by: breadth and depth of outcome measures and case-mix variables, repeated measurements over a 2-year period, and large cross-sectional (N-22,785) and longitudinal (N-2,546) samples. Of the longitudinal sample, 37% are 65 and older. The proposed research would extend MOS analyses, without additional data collection, to determine differences between health care delivery systems in physician style of caring for elderly and middle-aged patients, the characteristics of elderly patients and their physicians that account for variations in practice styles, and the implications for patient outcomes. Practice style is defined in terms of the quality of interpersonal aspects of the doctor-patient relationship, the extent of patient involvement in decision-making and information exchange, and the presence of counseling for mental health problems and personal health practices. Three health outcomes (clinical status, functioning in everyday activities, and general well-being) and patient satisfaction will be studied.