All-cause and cause-specific mortality among white U.S. men and women are analyzed using the NHANES I data (1971-1975) and epidemiologic follow-up to 1992, to examine the effect of physical stature on mortality, controlling for other confounding variables within a discrete-time framework. We find an association between mortality and both body mass index (BMI) and height, but the height effect is sensitive with respect to the age range under consideration. Although the resulting minimum-mortality BMI is higher than the widely accepted healthy range, the recent increase in weight implies that further gains in life expectancy are unlikely to derive from the anthropometry-mortality relationship.