The importance of changes in glucagon in the regulation of hepatic glucose production (Ra) during exercise has been questioned, as an increase in arterial immunoreactive glucagon (IRG) is not always detectable. However, IRG in the portal vein (PV) and not in the artery is most relevant, as flow through PV is approximately 80% of liver blood flow. To assess the extent that arterial IRG reflects the levels the liver is exposed to in PV, dogs (n = 5) were implanted with catheters in a carotid artery, hepatic vein (HV), and PV. Dogs were studied > or = 16 days later during rest and 150 min of moderate treadmill exercise, with indocyanine green and [3-3H]glucose infused to assess hepatic plasma flow (HPF) and hepatic Ra. IRG was 66 +/- 7, 73 +/- 8, and 81 +/- 7 pg/ml in the artery, HV, and PV at rest; it rose at 10 and 150 min of exercise to 89 +/- 9 and 127 +/- 13 pg/ml in the artery, 106 +/- 17 and 186 +/- 21 pg/ml in HV, and, by considerably more, to 153 +/- 20 and 261 +/- 25 pg/ml in PV. HPF fell by approximately 30% with exercise. The fall in HPF accounted for < 11% of the increased arterial-to-PV IRG gradient during exercise, with increased splanchnic IRG release comprising the remainder. Ra was linearly related to IRG levels in the three vessels.(ABSTRACT TRUNCATED AT 250 WORDS)