The pharmacokinetics of fusidic acid and 3-ketofusidic acid were investigated in cholestatic and noncholestatic patients after intravenous administration of single and multiple doses of 500 mg of sodium fusidate. The patients, all with low serum albumin levels, were divided into three groups. Group I consisted of six noncholestatic patients; group II consisted of nine mildly cholestatic patients with mild hepatic impairment (conjugated bilirubin, 47 mumol liter-1; alkaline phosphatase, 280 IU liter-1; gamma-glutamyltranspeptidase, 190 IU liter-1); group III consisted of six benign intrahepatic cholestatic patients with high isolated conjugated hyperbilirubinemia (98.1 mumol liter-1). Assays were performed by high-pressure liquid chromatography. At steady state, the mean peak concentrations in serum were 63.7, 44.9, and 92.2 micrograms ml-1 in groups I, II, and III, respectively; over a dosage interval, areas under the concentration-time curve were 411.1, 238.7, and 603.4 micrograms.h ml-1 and the mean body clearances were 0.34, 0.53, and 0.25 ml min kg-1 in groups I, II, and III, respectively. The accumulation ratio of fusidic acid increased from 2.8 and 2.4 in groups I and II to 4.2 in group III. At steady state, the ratios of the areas under the concentration-time curve from 0 to 8 h for 3-ketofusidic acid/fusidic acid were 0.11, 0.09, and 0.10 in the three groups, respectively. Only very small amounts of fusidic acid and 3-ketofusidic acid were found in urine. These results substantiate the following hypotheses. In group I and II patients the clearance is higher than that in healthy volunteers because of the increased free, unbound fraction of fusidic acid, a consequence of lower serum albumin concentrations, resulting in increased distribution in tissue and hepatic metabolism. In group III patients, the higher bilirubinemia results in competition with fusidic acid for the limited glucuronidation, thus compensating for the increased elimination with fusidic acid because of the low serum albumin concentration. These results suggest that fusidic acid can be administered normally even to patients with high bilirubinemia because the postoperative serum albumin concentration is usually low.