Aims: This study was aimed to characterize the population pharmacokinetics of $Arimidex^{(R)}$ (anastrozole) which has been administered to healthy Korean male volunteers. Methods: We have committed population pharmacokinetic analysis using $NONMEM^{(R)}$ VII with 408 plasma concentration data obtained from subjects who were administered with reference drug $Arimidex^{(R)}$ in a randomized, double-blind, two-way crossover bioequivalence trial conducted in Asan Medical Center. We estimated the parameters by first-order conditional estimation (FOCE) method, and then, observed the change of objective function values which had been calculated by using different compartment models, and explored the significant covariates. To evaluate the suitability and stability of final population model, we performed bootstrapping and visual/numerical predictive checks using Wings for NONMEM and Perl-speaks-NONMEM software. We simulated the plasma concentration corresponding to daily doses of anastrozole 1 mg through 10 days. Results: The plasma concentration data of $Arimidex^{(R)}$ is best fitted by first-order two-compartment linear model. The estimated absorption rate constant value is $0.23;h^{-1}$, and the final estimates for the volume of distribution of central and peripheral compartment is calculated as 5.12 L and 319 L, respectively. Conclusions: We successfully constructed the final population pharmacokinetic model of $Arimidex^{(R)}$ in this study, which could be useful for various purposes such as Monte-Carlo simulation for the plasma concentration-time profiles of $Arimidex^{(R)}$.