The objective of the study was to develop valid and reliable discriminative indices which measure patient and relative satisfaction with two closely related aspects of medical care: the level of care received and their involvement in decisions regarding care. We generated items by literature review and interviews with patients, relatives, and health care providers. In the final questionnaires, we included the items identified most frequently as sources of dissatisfaction and rated most important by 102 patients and 153 relatives. To measure reliability and validity we administered the instruments to 105 patients and 75 relatives of competent patients and 89 relatives of incompetent patients. We constructed three questionnaires: the Patient Satisfaction Index with 23 items, the Relative of Competent Patient Satisfaction Index with 34 items, and the Relative of Incompetent Patient Satisfaction Index with 29 items. We found mean scores of 75-80% of the maximum possible score, with a wide range of scores. The intraclass correlation for the instruments varied from 0.86 to 0.94. Correlations with global ratings were high (0.59-0.75) and similar to predictions. Correlations with caregiver ratings were lower than predicted (0.18-0.22). For both patients and relatives, our instruments discriminate between those with higher and lower satisfaction with the level of medical care and with their involvement with decision-making.