The purpose of this study was to determine the role of direct hepatic adrenergic stimulation in the control of endogenous glucose production (R(a)) during moderate exercise in poorly controlled alloxan-diabetic dogs. Chronically catheterized and instrumented (flow probes on hepatic artery and portal vein) dogs were made diabetic by administration of alloxan. Each study consisted of a 120-min equilibration, 30-min basal, 150-min moderate exercise, 30-min recovery, and 30-min blockade test period. Either vehicle (control; n = 6) or alpha (phentolamine)- and beta (propranolol)-adrenergic blockers (HAB; n = 6) were infused in the portal vein. In both groups, epinephrine (Epi) and norepinephrine (NE) were infused in the portal vein during the blockade test period to create suprapharmacological levels at the liver. Isotopic ([3-(3)H]glucose, [U-(14)C]alanine) and arteriovenous difference methods were used to assess hepatic function. Arterial plasma glucose was similar in controls (345 +/- 24 mg/dl) and HAB (336 +/- 23 mg/dl) and was unchanged by exercise. Basal arterial insulin was 5 +/- 1 mU/ml in controls and 4 +/- 1 mU/ml in HAB and fell by approximately 50% during exercise in both groups. Basal arterial glucagon was similar in controls (56 +/- 10 pg/ml) and HAB (55 +/- 7 pg/ml) and rose similarly, by approximately 1.4-fold, with exercise in both groups. Despite greater arterial Epi and NE levels in HAB compared with controls during the basal and exercise periods, exercise-induced increases in catecholamines from basal were similar in both groups. Gluconeogenic conversion from alanine and lactate and the intrahepatic efficiency of this process were increased by twofold during exercise in both groups. R(a) rose similarly by 2.9 +/- 0.7 and 2.7 +/- 1.0 mg. kg(-1). min(-1) at time = 150 min during exercise in controls and HAB. During the blockade test period, arterial plasma glucose and R(a) rose to 454 +/- 43 mg/dl and 11.3 mg. kg(-1). min(-1) in controls, respectively, but were essentially unchanged in HAB. The attenuated response to the blockade test in HAB substantiates the effectiveness of the hepatic adrenergic blockade. In conclusion, these results demonstrate that direct hepatic adrenergic stimulation does not play a role in the stimulation of R(a) during exercise in poorly controlled diabetes.