This study was aimed to perform clinical trial simulations to evaluate the dose reduction strategy of docetaxel for Japanese patients with liver dysfunction, which we previously proposed. For this purpose, a log-binominal regression (LBR) was performed for febrile neutropenia (FN) induced by docetaxel in these patients. A LBR analysis was conducted using clinical data from cancer patients treated with docetaxel and incorporated in the subsequent trial simulation. Virtual patients with liver dysfunction were randomly assigned to receive the Japanese standard dose (60 mg/m(2)) or reduced dose (40 or 50 mg/m(2)) of docetaxel. The primary endpoint was overall survival of the reduced dose to the standard dose. The secondary endpoint was the number of patients who experienced FN in response to the two treatment regimens. From the LBR analysis, the performance status and the area under the plasma concentration-time curve (AUC) were selected as covariates associated significantly (p<0.05) with FN occurrence. From the results of the present trial simulation, the median proportion of patients who experienced FN was decreased by about 20% in the reduced dose arm. Non-inferiority criteria, the reduced dose group to the standard dose group were met in 85.5% of the simulated clinical trials with a decrease in the FN frequency. In conclusion, clinical trial simulation models for the efficacy (survival) and toxicity (FN) was first performed in Japanese patients, and the feasibility of docetaxel therapy for liver-dysfunction patients under the dose reduction strategy was supported.