In the present study, we examined how the arterial insulin level alters the ␣-cell response to a fall in plasma glucose in the conscious overnight fasted dog. Each study consisted of an equilibration (؊140 to ؊40 min), a control (؊40 to 0 min), and a test period (0 to 180 min), during which BAY R 3401 (10 mg/kg), a glycogen phosphorylase inhibitor, was administered orally to decrease glucose output in each of four groups (n ‫؍‬ 5). In group 1, saline was infused. In group 2, insulin was infused peripherally (3.6 pmol ⅐ kg ؊1 ⅐ min ؊1 ), and the arterial plasma glucose level was clamped to the level seen in group 1. In group 3, saline was infused, and euglycemia was maintained. In group 4, insulin (3.6 pmol ⅐ kg ؊1 ⅐ min ؊1 ) was given, and euglycemia was maintained by glucose infusion. In group 1, drug administration decreased the arterial plasma glucose level (mmol/l) from 5.8 ؎ 0.2 (basal) to 5.2 ؎ 0.3 and 4.4 ؎ 0.3 by 30 and 90 min, respectively (P < 0.01). Arterial plasma insulin levels (pmol/l) and the hepatic portalarterial difference in plasma insulin (pmol/l) decreased (P < 0.01) from 78 ؎ 18 and 90 ؎ 24 to 24 ؎ 6 and 12 ؎ 6 over the first 30 min of the test period. The arterial glucagon levels (ng/l) and the hepatic portal-arterial difference in plasma glucagon (ng/l) rose from 43 ؎ 5 and 5 ؎ 2 to 51 ؎ 5 and 10 ؎ 5 by 30 min (P < 0.05) and to 79 ؎ 16 and 31 ؎ 15 (P < 0.05) by 90 min, respectively. In group 2, in response to insulin infusion, arterial insulin (pmol/l) was elevated from 48 ؎ 6 to 132 ؎ 6 to an average of 156 ؎ 6. The hepatic portal-arterial difference in plasma insulin was eliminated, indicating a complete inhibition of endogenous insulin release. The arterial glucagon level (ng/l) and the hepatic portalarterial difference in plasma glucagon (ng/l) did not rise significantly (40 ؎ 5 and 7 ؎ 4 at basal, 44 ؎ 4 and 9 ؎ 4 at 90 min, and 44 ؎ 8 and 15 ؎ 7 at 180 min). In group 3, when euglycemia was maintained, the insulin and glucagon levels and the hepatic portal-arterial difference remained constant. In group 4, the arterial plasma glucose level remained basal (5.9 ؎ 1.1 mmol/l) throughout, whereas insulin infusion increased the arterial insulin level to an average of 138 ؎ 6 pmol/l. The hepatic portal-arterial difference in plasma insulin was again eliminated. Arterial glucagon level (ng/l) and the hepatic portal-arterial difference in plasma glucagon (ng/l) did not change significantly (43 ؎ 2 and 9 ؎ 2 at basal, 39 ؎ 3 and 9 ؎ 2 at 90 min, and 37 ؎ 3 and 7 ؎ 2 at 180 min). Thus, a difference of ϳ120 pmol/l in arterial insulin completely abolished the response of the ␣-cell to mild hypoglycemia. Diabetes 51:3033-3042, 2002