Cefazolin, a time-dependent first-generation cephalosporin with non-linear binding to albumin, is widely recommended for antimicrobial prophylaxis during liver surgery to decrease the incidence of postoperative wound infections. The recommended protocol (2 g IV at anesthesia induction followed by 1 g 4 h later) is expected to maintain the free cefazolin concentration in exposed intratissular fluids above its minimal inhibitory concentration (MIC) for potentially encountered microorganisms, from skin incision to skin closure. Since this dosing protocol fails to take into account either of patients status (total body weight and renal function) or of surgical and anesthetic consequences (variations of cardiac output and regional blood flows, progressive decrease of plasma albumin concentration) on cefazolin tissular pharmacokinetics, a physiological modeling study was conducted to investigate protocol suitability for liver surgery in six populations: obese (body mass index >34), renal insufficiency (GFR = 10, 30 or 50 ml min) and high intraoperative blood loss (three times that usually observed during this surgery) and none of these features referred to as controls. A previously validated physiologically based pharmacokinetic (PB-PK) model for cefazolin in humans was used and then further adapted to simulate obese or renal insufficiency patients as well as the consequences of general anesthesia and liver surgery on cefazolin pharmacokinetics. Clinical data required for simulation (intraoperative kinetics of percent expired isoflurane and plasma albumin concentration, mean intraoperative blood loss) were obtained from 10 patients who underwent right hepatectomy in our institution. Using a fixed MIC of 2 μg ml against potentially encountered bacteria, it was concluded that the recommended dosing schedule was suitable in all tested populations, including obese patients, although prolongation of the interval between injections appeared advisable for renal insufficiency patients. Furthermore, when a MIC of 3 μg ml was considered, the recommended cefazolin-dosing regimen failed to maintain sufficient free cefazolin concentrations in the interstitial fluids during surgery in all tested populations except renal insufficiency patients (GFR < 50 ml min).