A method for predicting the interindividual variability of human exposure for CYP3A4 substrates using Monte Carlo simulation was developed based on relevant factors. The coefficient of variation (CV) values for CYP3A4 content in human liver microsomes, hepatic blood flow, liver volume and body weight, and the unbound blood fraction were collected from the published literature. The parallel tube and dispersion models were found to be appropriate mathematical models to describe the pharmacokinetics (PK). Simulation results using 33% as the CV for CYP3A4 content reflected reported CV values of the area under the curve (AUC) for 40 CYP3A4 substrates for both intravenous and oral administration. We also successfully predicted the clearance of midazolam in Japanese and in European American subjects. In all cases, the simulated mean and SD values reflected the reported values. Thus, the interindividual variability of the AUC of CYP3A4 substrates was predictable for both intravenous and oral administration.