OBJECTIVE: The effectiveness of an androgenic nutritional supplement designed to enhance serum testosterone concentrations and prevent the formation of dihydrotestosterone and estrogen was investigated in healthy 3 to 58 year old men. DESIGN: Subjects were randomly assigned to consume a nutritional supplement (AND-HB) containing 300-mg androstenediol, 480-mg saw palmetto, 450-mg indole-3-carbinol, 300-mg chrysin, 1,500 mg gamma-linolenic acid and 1.350-mg Tribulus terrestris per day (n = 28), or placebo (n = 27) for 28 days. Subjects were stratified into age groups to represent the fourth (30 year olds, n = 20), fifth (40 year olds, n = 20) and sixth (50 year olds, n = 16) decades of life. MEASUREMENTS: Serum free testosterone, total testosterone, androstenedione, dihydrotestosterone, estradiol, prostate specific antigen and lipid concentrations were measured before supplementation and weekly for four weeks. RESULTS: Basal serum total testosterone, estradiol, and prostate specific antigen (PSA) concentrations were not different between age groups. Basal serum free testosterone concentrations were higher (p < 0.05) in the 30- (70.5 +/- 3.6 pmol/L) than in the 50 year olds (50.8 +/- 4.5 pmol/L). Basal serum androstenedione and dihydrotestosterone (DHT) concentrations were significantly higher in the 30- (for androstenedione and DHT, respectively, 10.4 +/- 0.6 nmol/L and 2198.2 +/- 166.5 pmol/L) than in the 40- (6.8 +/- 0.5 nmol/L and 1736.8 +/- 152.0 pmol/L) or 50 year olds (6.0 +/- 0.7 nmol/L and 1983.7 +/- 147.8 pmol/L). Basal serum hormone concentrations did not differ between the treatment groups. Serum concentrations of total testosterone and PSA were unchanged by supplementation. Ingestion of AND-HB resulted in increased (p < 0.05) serum androstenedione (174%), free testosterone (37%), DHT (57%) and estradiol (86%) throughout the four weeks. There was no relationship between the increases in serum free testosterone, androstenedione, DHT, or estradiol and age (r2 = 0.08, 0.03, 0.05 and 0.02, respectively). Serum HDL-C concentrations were reduced (p < 0.05) by 0.14 mmol/L in AND-HB. CONCLUSIONS: These data indicate that ingestion of androstenediol combined with herbal products does not prevent the formation of estradiol and dihydrotestosterone.