Recent data suggest that insulin is a modulator of ovarian and adrenal steroidogenesis and that, in the ovary of hyperandrogenic women, hyperinsulinemia might cause dysregulation of cytochrome P450c17 alpha activity. To further assess in vivo the effects of insulin on adrenal steroidogenesis, ACTH stimulation was carried out in 21 hyperandrogenic women during a 3-h hyperinsulinemic (80 mU/m2-min) euglycemic clamp. In all of these women the procedure was repeated during saline infusion as n control. In nonamenorrheic patients, the tests were performed in the early follicular phase of two different menstrual cycles. Serum cortisol, progesterone, 17-hydroxypregnenolone (17-OHJPREG). 17-hydroxyprogesterone (17-OHP), dehydroepiandrosterone (DHEA), and androstenedione (A) were measured after 2 h of insulin or saline infusion (zero time) and, subsequently, 30 and 60 min after an iv bolus of 0.25 mg ACTH-(1-24). At zero time, no difference was found in the serum steroid concentrations between the two protocols. ACTH-stimulated serum 17-OHPREG and, to a lesser extent, 17-OHP were significantly higher during insulin than during saline infusion (peaks, 60.6 +/- 9.0 vs. 40.7 +/- 7.9 and 7.7 +/- 7.7 vs. 6.6 +/- 0.6 nmol/L; P < 0.005 and P < 0.01, respectively). Serum DHEA was also slightly higher during hyperinsulinemia, although only after 30 min (54.5 +/- 3.0 vs. 48.2 +/- 4.2 nmol/L; P < 0.05). No statistically significant difference in the cortisol, progesterone, or androstenedione response to ACTH was found between the two protocols. ACTH-stimulated 17-OHPREG/DHEA and 17-OHP/A molar ratios, indexes of apparent 17,20-lyase activity, were significantly higher during the clamp studies than during saline infusion (by ANOVA, F = 12.8; P < 0.001 and F = 6.7; P < 0.005, respectively), suggesting an impaired enzyme activity. These in vivo data support the hypothesis that insulin potentiates ACTH-stimulated steroidogenesis. This effect of insulin seems to be associated with a relative impairment of 17,20-lyase activity.