Oral glucose administration increases insulin secretion to a greater extent than peripheral glucose infusion (incretin effect). It also augments protal vein blood flow, hepatic uptake of glucose, and fractional hepatic extraction of insulin. The mechanisms for these various effects are not known but could involve both neurogenic stimuli and gut hormones. The present studies examined the effect of a non-nutrient drink, 1 g/kg body wt oral mannitol, on these parameters during an intravenous glucose infusion in conscious dogs. The dogs had chronically implanted Doppler flow probes on the portal vein and hepatic artery and catheters in the portal vein, hepatic vein, and femoral artery. After a 30-min control period, an infusion of atropine, propranolol, phentolamine, or propranolol and phentolamine was begun. Thirty minutes later, glucose (13 mg.kg-1.min-1) was then infused into a peripheral vein for 120 min with continuation of the atropine and adrenergic blockade. Water or mannitol (10% solution) was administered orally 50 min after the initiation of the glucose infusion. Mannitol, but not water, significantly enhanced the insulin response to intravenous glucose, as indicated by higher insulin concentrations in the portal vein as well as more rapid reduction of the plasma glucose. This incretin effect was significantly attenuated by infusion of propranolol but not by atropine or phentolamine. Mannitol did not increase portal vein blood flow or have any effect on the hepatic uptake of glucose or the fractional hepatic extraction of insulin. Thus absorption of nutrient is not necessary for the incretin effect but is for the increased portal vein blood flow and increased fractional extraction of insulin.(ABSTRACT TRUNCATED AT 250 WORDS)