Four beef steers (average BW, 246 kg) were used in a split-plot design with five bovine insulin (I) infusion rates (5, 10, 20, 40, and 80 mU.kg-1.h-1) in the main plot and two infusion sites, mesenteric (M) and jugular (J) veins, in the subplot. Steers were fed a medium-energy diet at .235 Mcal of ME/kg BW.75 daily in 12 equal feedings at 2-h intervals. Catheters were placed in the mesenteric, hepatic-portal, and hepatic veins and in the abdominal aorta. Blood was sampled from the arterial, portal, and hepatic catheters at 20-min intervals for 1 h before I infusion. Glucose was infused intrajugularly to maintain euglycemia during the I infusion, with arterial glucose monitored at 10- to 15-min intervals. After at least 2.5 h, blood was again sampled at 20-min intervals for 1 h. Blood flow was determined by downstream dilution of p-aminohippurate. Arterial I concentrations (+/- SE) at the greatest I infusion rates were 183.5 +/- 10.46 (J) and 179.0 +/- 6.64 (M) microU/mL. Portal I concentration tended to be greater during M than during J infusion (e.g., J, 199.9 +/- 10.48 vs M, 225.8 +/- 8.99 microU/mL at the greatest dose). Hepatic glucose production at the larger three I doses reached a plateau near 40% of the preinfusion production rate (.57 +/- .02 mmol.kg-1.h-1 vs J, .23 +/- .029 and M, 27 +/- .037). Urea N concentration decreased, but portal uptake or hepatic release of urea N was largely unaffected by I dose or site of infusion.(ABSTRACT TRUNCATED AT 250 WORDS)