To delineate the role of the route of glucose administration on liver glycogen synthesis, we administered [1-13C]glucose either by intravenous or intraduodenal infusion to chronically catheterized 24-h fasted rats and performed 1) intravenous-hyperglycemic clamp (group I, n = 9), portal vein plasma glucose and insulin concentrations were 216 +/- 6 mg/dl and 23.9 +/- 4.2 ng/ml; 2) intraduodenal-hyperglycemic infusion (group II, n = 8), portal vein glucose and insulin concentrations were 219 +/- 6 mg/dl and 17.5 +/- 2.7 ng/ml; and 3) intravenous-hyperglycemic-suprahyperinsulinemic clamp (group III, n = 5), portal vein glucose and insulin concentrations were 203 +/- 12 mg/dl and 44.6 +/- 5.0 ng/ml. The mean glucose infusion rates (mumol.kg-1.min-1) and glycogenic rates (mumol.g liver-1.min-1) were 201 +/- 8, 0.34 +/- 0.05; 129 +/- 3, 0.73 +/- 0.11; and 269 +/- 19, 0.38 +/- 0.08 in groups I-III, respectively. The percent of glycogen synthesized by the direct pathway was group I = 43 +/- 6%, group II = 44 +/- 6%, and group III = 46 +/- 7%. In conclusion, despite similar or lower portal vein insulin and glucose concentrations, the intraduodenal route of glucose administration (group II), compared with the intravenous route (groups I and III), markedly increased the total amount of liver glycogen synthesized without altering the percent of the direct vs. indirect pathways by which liver glycogen was repleted.