Although many older persons choose to postpone retirement, the health consequences of this decision are still debated. This study aimed at determining the impact of continued employment on life-style, well-being, health and survival in a cohort of community-dwelling 70-year-olds. A longitudinal study of a homogeneously aged cohort was conducted in two stages. Data from extensive interviews, physical examinations and clinical laboratory information were obtained from west Jerusalem residents born in 1920-1 through home-based examinations. In 1990-1, 162 women and 218 men were examined, at age 70, and in 1997-8, 390 women and 398 men at age 77. Two hundred and thirty one subjects participated in both study stages and were available for longitudinal analysis. For each subject at each study stage, a comprehensive profile detailed social characteristics, including life-style and personal functions, and medical status including illnesses and use of medical services. At each stage, employed subjects were compared with those who neither worked for compensation nor as volunteers. Logistic regression compared health outcomes at age 77 for subjects who worked at age 70 with those of subjects who had not. At age 70, 37 women and 98 men were gainfully employed, whereas 125 women and 120 men did not work. At age 77, wage-earners numbered 26 women and 95 men, and non-workers 364 women and 303 men. In the cross-sectional analyses at ages 70 and 77, work correlated with better perceived health and greater independence for both women and men. Independent of pre-existing economic difficulties, education, illnesses, functional dependence and self-assessed health, workers at age 70 after seven years were in better health (OR 2.12, 95% CI 1.04, 4.30), had more ADL independence, (OR 2.60, 95% CI 1.08, 6.29) and increased survival rates (OR 2.13, 95% CI 1.06, 4.28). Independent of baseline status, working at age 70 correlates with health, self-sufficiency and longevity.