GLP-1 administration decreases blood glucose levels in normal subjects and non-insulin-dependent diabetes mellitus patients and is therefore proposed as a treatment for diabetic hyperglycaemia. The glucose lowering effect of GLP-1 is glucose dependent and therefore self-limiting, but it is not known to which extent counterregulatory mechanisms participate in this. GLP-1 was infused i.v. into 8 healthy subjects after an overnight fast at a rate of 100 pmol kg-1 h-1 for 1 h with and without beta-adrenoceptor blockade (i.v. bolus of 5 mg propranolol followed by a continuous infusion of 0.08 mg min-1). In a control experiment, saline and propranolol were infused. Hepatic glucose production was measured and blood was analysed for plasma glucose, insulin, glucagon, catecholamines, and radioactivity. Plasma GLP-1 levels were similar on the two GLP-1 infusion days and resulted in: (1) a significant decrease in plasma glucose from 5.2 +/- 0.2 to 4.1 +/- 0.1 mmol l-1 with GLP-1/propranolol infusion, and from 5.2 +/- 0.1 to 4.0 +/- 0.1 mmol l-1 with GLP-1/saline infusion (NS); (2) a corresponding significant increase in plasma insulin from 58.0 +/- 6.3 to 144.5 +/- 22.3 pmol l-1 and from 61.7 +/- 6.4 to 148.2 +/- 34.0 pmol l-1, respectively (NS); (3) a significant decrease in plasma glucagon from 11.7 +/- 1.6 to 6.5 +/- 1.5 pmol l-1 and from 10.4 +/- 1.6 to 4.6 +/- 1.0 pmol l-1, respectively; (4) a significant decrease in the rate of glucose appearance which was not significantly different on the two GLP-1 infusion days; and (5) an increase in catecholamine levels in the GLP-1/saline experiment and also in the beta-blockade experiments. We conclude that adrenergic counterregulation plays an insignificant role in curtailing GLP-1's glucose lowering effect.