In a 25-year follow-up of 2370 middle-aged civil servants and spouses of civil servants, the authors studied the effect of the number of parents (none, one, or both) which a middle-aged person still has alive on this person's survival. Among men, no relationship was found with the number of parents alive after 15 years of follow-up and a weak gradient was found after 25 years. Among women, there was a clear gradient of mortality according to the number of parents alive, both after 15 and after 25 years. In a logistic regression analysis of mortality, after 25 years, the age-adjusted effect of having both parents alive relative to none was significant in men and women (for men: odds ratio = 0.63, 95% confidence interval = 0.42-0.97; for women: odds ratio = 0.36, 95% confidence interval = 0.19-0.69). These estimates did not change materially upon inclusion of other determinants for chronic disease and death into the model (smoking, blood pressure, serum cholesterol, and body mass index). The authors concluded that the parental effect was independent of these risk factors.