To assess the effects of 16 weeks of heavy resistance exercise training (RE) on insulin sensitivity and secretion in healthy older men aged 64 to 75 years (N = 15), stable-label ([6,6,2H2]glucose) intravenous glucose tolerance tests (IVGTTs) were performed before and 7 days after the last bout of exercise. Glucose disappearance rate (Rd) and an index of insulin sensitivity (Si*) were derived using the minimal model of labeled glucose disappearance, and insulin secretion parameters were derived from C-peptide and glucose concentrations measured during the IVGTT, using a minimal model of C-peptide secretion and kinetics. Each subject trained at an intensity of 70% to 95% maximum strength 4 d/wk for 16 weeks on Nautilus (DeLand, FL) weight-training equipment. In conjunction with exercise, six men received daily injections of recombinant human growth hormone ([rhGH] 12.5 to 24 microg/kg/d) and the other nine received placebo injections. GH/placebo injections were administered in a double-blind randomized fashion. The RE program was supervised and progressive in nature, consisting of both upper-and lower-body exercises, and significantly increased muscle strength (P < .05) with no additional benefit from rhGH except for a tendency toward a greater increase in fat-free mass (FFM) in the RE + GH group (P = .06). Peak glucose Rd increased following RE (P < 01), and there was a trend for an improved Si* (ie, from 6.79 +/- 1.14 to 8.42 +/- 0.89 x 10(4) per min/[microU/mL], P = .06). Peak glucose Rd and Si* were unchanged in the RE + GH group following treatment. First- and second-phase insulin secretion were not affected by RE or RE + GH. Glucose tolerance, quantified as the glucose disappearance constant (Kg) between 10 and 32 minutes of the IVGTT, was unchanged by exercise or hormone treatment. These findings support those of a recent study that used the hyperinsulinemic-euglycemic clamp technique (Miller et al, J Appl Physiol 77:1122-1127, 1994), and suggest that when healthy older men engage in RE, whole-body glucose Rd and Si* are improved, and these beneficial effects are not only due to the acute effects of the last bout of exercise. Additionally, in six subjects who received GH, glucose Rd and Si* were not significantly improved following the RE program. Although this may suggest that GH can diminish improvements in glucose Rd and Si* that result from RE, further study is needed to confirm this observation.