We previously demonstrated that during a 10-day fast in mildly obese men, urinary gonadotropin excretion significantly increased, and serum testosterone concentrations significantly decreased. The mechanisms by which these changes occur are unknown. We postulated that the mechanism of the gonadotropinuria might involve decreased proximal renal tubular reabsorption of gonadotropins during fasting and might be related to renal tubular reabsorption of ketones during fasting, a process that is enhanced by carbohydrate (CHO) administration. We studied the effects of CHO supplementation on ketosis, ketonuria, and reproductive hormone secretion and excretion in 14 mildly obese men, 24-54 yr old, who were 14-69% above ideal body weight. Group I (n = 6) received no CHO supplementation, group II (n = 4) received 15 g CHO, and group III (n = 4) received 45 g CHO daily during the 10-day fast (F). During the control (C) and refeeding (R) periods, all subjects received a 1500-cal diet. Daily 24-h urine collections were made for the measurement of total ketones (millimolar concentrations) and LH and FSH (expressed as international units of the Second International Reference Preparation of human menopausal gonadotropin). Values (mean +/- SE) for 3 representative days (control day 3, fasting day 8, and refeeding day 3) for all subjects are shown below: (table; see text) We also studied the effects of CHO supplementation on serum levels of pituitary gonadotropins, LH and FSH responses to exogenous LHRH stimulation, biological activity of LH, and circulating total and free testosterone levels. Neither dose of CHO prevented the decline in total and free testosterone levels. Serum LH concentrations, as measured by both the RIA and in vitro bioassay did not change significantly with fasting. Serum FSH concentrations in daily samples did not change significantly. The previously reported decline in the FSH response to LHRH stimulation with fasting was not prevented by CHO. We conclude that CHO supplementation prevents the gonadotropinuria of fasting in men. The effect appears to occur in the kidney. The mechanisms may be related to that by which CHO promotes the renal tubular reabsorption of ketones. The reduced serum testosterone level cannot be explained by a lack of biologically active LH. It appears that fasting has a direct effect on the testis, possibly by reducing its responsiveness to gonadotropic stimulation or by inhibiting steroidogenesis.