Plasma clearance of D-sorbitol, a nontoxic polyol, occurs predominantly in the liver and has been used to measure functional liver blood flow after bolus and steady- state intravenous administration. However, it is not known which of these two administration methods is superior. Therefore, plasma D-sorbitol clearance was studied in an animal model both after a bolus dose and under steady-state (SS) conditions and compared directly with liver blood flow, under normal conditions, and after the induction of endotoxin (LPS) sepsis. Adult male Wistar rats (526 +/- 38 g body wt; n = 27) were anesthetized and mechanically ventilated. Hemodynamics, hepatic arterial flow, and portal venous flow were measured. Two groups were studied, namely healthy animals that served as controls and a sepsis group that received 5 mg/kg LPS intravenously (Escherichia coli O127:B8). Each animal received either a SS infusion (0.1 mg/100 g body wt per min) or a bolus (3 mg/100 g body wt) of a 5% D-sorbitol solution intravenously in a randomized order. After the initial measurements and a 60-min pause time in between (T(1/2,sorbitol) = 9 min), a crossover was done. The hepatic clearance of D-sorbitol in the control group showed a good correlation between bolus and SS (Spearman's r = 0.7681, P = 0.0004), and both techniques correlated well with total liver blood flow (TLBF) (r = 0.7239, P = 0.0023 and r = 0.7226, P = 0.0023, respectively). Also in the sepsis group there was a good correlation between bolus and SS sorbitol clearance (r = 0.6655, P = 0.0182). In the sepsis group, only the SS clearance correlated with TLBF (r = 0.6434, P = 0.024). In conclusion, in normal and under septic conditions, hepatic clearance of D-sorbitol either by bolus or a SS infusion is comparable. In healthy animals, this also correlated well with TLBF but not in septic conditions. However, this is expected because of the changes in the liver microcirculation, shunting, and decreased hepatocyte function in sepsis.