Data from a 20-year longitudinal study of persons over 60 were analyzed by step-wise multiple regression to test for declines in function with age, for terminal decline (linear relationship to time before death), and for terminal drop (curvilinear relationship to time before death). There were no substanial terminal drop effects. All health measures had substantial declines with age, and the physician's physical function rating, as well as the individual's self-rating, showed additional terminal decline. Intelligence had a substantial decline with age and a small terminal decline. Most activities declined with age but had no significant terminal decline. Despite these declines in health, intelligence, and activities, there was little or no decline in satisfaction or adjustment.