Etoposide, an oil anticancer agent, has shown problematic interpatient variability in absorption in the intestine as well as excretion in the liver. Bone marrow suppression occurs easily in the case of increased plasma free etoposide fraction. In this study, we investigated a possible method for the pharmacokinetic estimation of free plasma etoposide using serum markers; such as, fasting serum total bile acid (TBA) and indocyanine green (ICG). The subjects for this study included nine patients with hepatocellular carcinoma complicated by anicteric liver cirrhosis (LC-HCC), five cancer patients (except HCC) without liver dysfunction. Oral dosage of etoposide was 50 mg kg - 1 body weight day - 1 . Total etoposide was measured using ultrasensitive HPLC, and protein-unbound etoposide was determined using the same HPLC method after ultrafiltration. In the LC-HCC group, maximum free etoposide level was detected 4 h after administration in all patients. The free ratio after 4 h was 12.97% (mean). On the other hand, in cancer patients (except HCC) with normal liver function, almost all the patients (four of five) had no free etoposide. We have developed a formula which estimates the protein unbinding ratio 4 h after administration (%) (Ratio-4 h). Using regression analysis, ICG, TBA, and platelet were found to be important factors, and the following formula was obtained using stepwise regression analysis: Predicted value (Ratio-4 h) = 0.001 TBA + 0.004 ICG + 0.011 (R 2 = 0.811, P < 0.001). In conclusion, it was demonstrated that serum fasting TBA and ICG could be significant markers for the evaluation of etoposide pharmacokinetics, especially the protein unbinding ratio.