The role which glucagon could play in the mechanism of fasting natriuresis and renal sodium retention associated with carbohydrate refeeding was studied in thirty-seven non-diabetic obese subjects. In nine obese subjects undergoing a 7 day fast without any additional treatment (control group), the renal sodium excretion exceeded intake through the whole experimental period, with maximal natriuresis on day 2 of the fast. Blood glucose and plasma insulin (IRI) levels fell rapidly from the first day of fast on, while pancreatic glucagon (IRG) titres rose from day 1 to day 4, declining slightly thereafter. When additional subjects received intravenous glucose on day 4 (n = 6), there was a rise in blood glucose concentration and in IRI associated with a rapid drop in IRG restricted to the period of glucose infusion. The resulting antinatriuresis occurred essentially during the following 36 h, while IRG and IRI levels had returned to fasting levels. A comparable glucose load on day 4 associated with 0.1 mg glucagon (n = 5) still led to the glucose-induced antinatriuresis while 1 mg glucagon added to a similar glucose infusion completely abolished its antinatriuretic effect (n = 6). Glucagon infused alone on day 4 of fast aggravated fasting natriuresis (n = 5) but was devoid of this effect when administered 24 h after the glucose load (n = 6). These data indicate that fasting hyperglucagonaemia or its reduction upon glucose refeeding, cannot be considered as directly involved in renal mechanism(s) responsible for fasting natriuresis of antinatriuretic effects of carbohydrate. It is suggested that the role of glucagon is indirect, possibly through its influence on ketogenesis which in turn may alter renal sodium handling.