BACKGROUND Changing plasma protein concentrations may affect the protein binding and pharmacokinetics of drugs in the postoperative phase. Therefore, the authors evaluated the pharmacokinetics of ropivacaine, administered by 72-h epidural infusion to provide postoperative analgesia. METHODS Twenty-eight patients, scheduled for major orthopedic surgery during combined epidural and general anesthesia received a bolus dose of ropivacaine (50 or 75 mg), followed by constant-rate (10 ml/h) epidural infusion of ropivacaine 2 mg/ml (group 1) or 3 mg/ml (group 2). Total and unbound plasma concentrations of ropivacaine and pipecoloxylidide and plasma concentrations of alpha1-acid glycoprotein were determined. In addition, the urinary excretion of ropivacaine and major metabolites was measured. RESULTS Total plasma concentrations of ropivacaine increased steadily during the infusion, reaching 2.7 +/- 0.7 and 2.9 +/- 0.5 mg/l in groups 1 and 2 after 72 h constant-rate infusion. Unbound ropivacaine concentrations reached average steady state levels of approximately 0.06 and 0.07 mg/l. Total and unbound concentrations of pipecoloxylidide increased to 1.0 +/- 0.4 and 0.4 +/- 0.2 mg/l (group 1) and 1.2 +/- 0.4 and 0.5 +/- 0.1 mg/l (group 2) after 72 h infusion. alpha1-Acid glycoprotein concentrations initially decreased, but thereafter increased steadily to approximately twice the baseline values. CONCLUSIONS Postoperative increases in plasma alpha1-acid glycoprotein concentrations enhance the protein binding of ropivacaine and pipecoloxylidide, causing divergence of total and unbound plasma concentrations.