It has been proposed that the decline in glucose tolerance with age is not a primary aging effect but is secondary to a combination of other age-associated characteristics, i.e., disease, medication, obesity, central and upper-body fat deposition, and inactivity. To test this hypothesis, we first eliminated from analysis the Baltimore Longitudinal Study of Aging participants with identifiable diseases or medications known to influence glucose tolerance. Seven hundred forty-three men and women, aged 17–92 yr, remained for analysis. As indices of fatness, body mass index and percent body fat were determined. As indices of body fat distribution, waist-hip ratio and subscapular triceps skin-fold ratio were calculated. As indices of fitness, physical activity level, determined by detailed questionnaire, and maximum 02 consumption were calculated. We tested whether the effect of age on glucose tolerance remains when data were adjusted for fatness, fitness, and fat distribution; 2-h glucose values were 6.61, 6.78, and 7.83 mM for young (17–39 yr), middle-aged (40–59 yr), and old (60–92 yr) men and 6.22, 6.22, and 7.28 mM for the three groups of women, respectively. The differences between the young and middle-aged groups were not significant, but the old groups had significantly higher values than young or middle-aged groups. Fatness, fitness, and fat distribution can account for the decline in glucose tolerance from the young adult to the middle-aged years. However, age remains a significant determinant of the further decline in glucose tolerance of healthy old subjects.