Major liver resections are associated with considerable morbidity and mortality. Gut-derived bacteria and bacterial endotoxin (LPS) are considered to play a central role in the pathophysiology of these complications. Like human BPI, rBPI(21) binds to LPS from Gram-negative bacteria. By binding and clearing of LPS, rBPI(21) can inhibit a number of endotoxin-induced humoral and cellular responses. Because of this capacity, rBPI(21) could partially compensate for the loss of hepatic mononuclear phagocytic system function after liver resection. However, the liver is also thought to be an important organ for the clearance of BPI, and reduction of liver mass could result in a decreased clearance and exceedingly high plasma levels of rBPI(21). In this study we therefore investigated the pharmacokinetics of rBPI(21) in rats and in patients undergoing a major liver resection. Rats were administered an intravenous (i.v.) bolus of rBPI(21) after undergoing a 60% or 80% hepatectomy (with sham-operated controls). Patients undergoing a hemihepatectomy and healthy volunteers received rBPI(21) or placebo by continuous i,v. infusion for 48 h. Plasma concentrations were measured by sandwich ELISA. In rats, 60% hepatectomy did not consistently change the clearance of rBPI(21), whereas 80% hepatectomy decreased the clearance of rBPI(21) severalfold. In hemihepatectomized patients, the clearance of rBPI(21) after major hepatectomy was also slower, when compared with healthy volunteers, but this difference had disappeared within 24 h, Our data indicate that the administration of rBPI(21) in patients undergoing liver resection is well tolerated and does not result in exceedingly high plasma levels. Additional studies on the efficacy of rBPI(21) in the prevention of complications after hepatectomy are needed.