We previously reported that infection decreases hepatic glucose uptake when glucose is given as a constant peripheral glucose infusion (8 mg. kg(-1) x min(-1)). This impairment persisted despite greater hyperinsulinemia in the infected group. In a normal setting, hepatic glucose uptake can be further enhanced if glucose is given gastrointestinally. Thus the aim of this study was to determine whether hepatic glucose uptake is impaired during an infection when glucose is given gastrointestinally. Thirty-six hours before study, a sham (SH, n = 7) or Escherichia coli-containing (2 x 10(9) organisms/kg; INF; n = 7) fibrin clot was placed in the peritoneal cavity of chronically catheterized dogs. After the 36 h, a glucose bolus (150 mg/kg) followed by a continuous infusion (8 mg. kg(-1). min(-1)) of glucose was given intraduodenally to conscious dogs for 240 min. Tracer ([3-(3)H]glucose and [U-(14)C]glucose) and arterial-venous difference techniques were used to assess hepatic and intestinal glucose metabolism. Infection increased hepatic blood flow (35 +/- 5 vs. 47+/-3 ml x g(-1) x min(-1); SH vs. INF) and basal glucose rate of appearance (2.1+/-0.2 vs. 3.3+/-0.1 mg x kg(-1) x min(-1)). Arterial insulin concentrations increased similarly in SH and INF during the last hour of glucose infusion (38+/-8 vs. 46+/-20 microU/ml), and arterial glucagon concentrations fell (62+/-14 to 30+/-3 vs. 624+/-191 to 208+/-97 pg/ml). Net intestinal glucose absorption was decreased in INF, attenuating the increase in blood glucose caused by the glucose load. Despite this, net hepatic glucose uptake (1.6+/-0.8 vs. 2.4+/- 0.9 mg x kg(-1) x min(-1); SH vs. INF) and consequently tracer-determined glycogen synthesis (1.3+/-0.3 vs. 1.0+/-0.3 mg. kg(-1) x min(-1)) were similar between groups. In summary, infection impairs net glucose absorption, but not net hepatic glucose uptake or glycogen deposition, when glucose is given intraduodenally.