The insulin effect, evaluated with the euglycaemic clamp technique, was studied before and after hypoglycaemia in 7 patients with Type 1 (insulin-dependent) diabetes. Following an initial 2 h clamp (clamp I) hypoglycaemia was induced and 2 h later a second clamp (clamp II), identical to the former, was performed. Each subject was studied twice; during infusion with saline (placebo) or propranolol. Glucose production and disposal were studied with the 3(3H)glucose technique. During placebo infusion, hypoglycaemia elicited an insulin resistance leading to approx. 50% reduction in the steady state glucose infusion rate during clamp II as compared to clamp I (clamp I 2.58 +/- 0.32, clamp II 1.26 +/- 0.08 mg . kg-1 . min-1, p less than 0.02). The insulin resistance was prevented by infusing propranolol (clamp I 2.29 +/- 0.29, clamp II 2.85 +/- 0.56 mg . kg-1 . min-1). The posthypoglycaemic insulin resistance was due to a less pronounced insulin effect on both glucose production (clamp I 0.29 +/- 0.21, clamp II 0.86 +/- 0.19 mg . kg-1 . min-1, p less than 0.05) and glucose utilisation (clamp I 2.84 +/- 0.26, clamp II 2.13 +/- 0.23 mg . kg-1 . min-1, p less than 0.05). The insulin resistance on both glucose production and utilisation was prevented by propranolol. Thus, the present study demonstrates that hypoglycaemia elicits a prolonged insulin resistance which is due to a less pronounced effect of insulin to both inhibit splanchnic glucose production and to stimulate peripheral glucose utilisation. The insulin resistance is due to beta-adrenergic stimulation and can be prevented by propranolol.