The relationship between insulin resistance and aging is still debated. This study aims to investigate the role that age-related differences in plasma dehydroepiandrosterone sulfate (DHEAS) concentration may have on insulin action. For this reason, 75 subjects (42 men and 33 women) with a wide age range (21 to 106 years) were studied. In all subjects, plasma DHEAS and total testosterone concentrations were measured and a euglycemic clamp was used, but substrate oxidation was not determined in centenarians (n = 15). Plasma DHEAS correlated with age (r = -.77, P < .001) and whole-body glucose disposal (WBGD) (r = .57, P < .001). After controlling for age, sex, body fat, and waist to hip ratio (WHR), the association between plasma DHEAS and WBGD was still observed (r = .31, P < .005). Comparing subjects at the third tertile versus those at the first and second tertiles of plasma age-adjusted DHEAS concentration, the former group showed a weaker association between WBGD and age (r = -.38, P < .05) than the latter group (r = -.43, P < .002). The difference between the two regression lines was also significant (P < .03). After controlling for sex, body fat, and WHR, the association between plasma DHEAS and WBGD was dependent on the age of the subjects, being strong in adults (n = 30, age < 50 years, r = .69, P < .001), weak in old subjects (n = 30, age 51 to 99 years, r = .23, P < .05), and absent in centenarians (r = -.05, P < .88). With the subjects divided by sex throughout the different age groups, the univariate association between plasma DHEAS and WBGD was present in females (r = .43, P < .01) but not in males (r = .17, P < .32). Plasma total testosterone and insulin-like growth factor-1 (IGF-1) concentrations declined with advancing age and were significantly correlated with DHEAS and WBGD. In a multivariate analysis with WBGD as the dependent variable, a model including age, sex, body fat, WHR, DHEAS, total testosterone, and IGF-1 explained 66% of WBGD variability, with DHEAS significantly and independently associated with WBGD (P < .004). In conclusion, the negative relationship between advancing age and insulin action seems related to plasma DHEAS concentration. Differences in plasma total testosterone and IGF-1 concentrations may provide a further contribution to the relationship between DHEAS and WBGD.