Background and Aim: Hepatic enzyme CYP2D6 plays an important role in metabolizing many drugs and its genetic polymorphism results in significant individual therapeutic difference.Estimating the phenotype of CYP2D6 before the administration of its substrate will be of great benefit to individual therapy.Determining metabolic ratio (MR) from single-point plasma is potentially a good phenotyping method of CYP2D6 to reduce the required time interval and increase the reliability of data.It is difficult to conduct large sample size clinical trials to evaluate this phenotyping method for multiple plasma points, given that many blood samples are taken from each subject.A physiologically-based pharmacokinetic (PBPK) model can be developed to do simulations based on the large virtual Chinese population to evaluate plasma phenotyping methods of CYP2D6, to optimize the trial designs for subsequent studies.Methods: PBPKmodels of dextromethorphan (DM) and its metabolite dextrorphan (DX) were developed in extensive metabolizer (EM), intermediate metabolizer (IM), and poor metabolizer (PM) groups, respectively, in Simcyp software.The simulation data were based on 1000 virtual Chinese subjects with reported frequencies of phenotypes.The simulation data were used to evaluate the phenotyping methods of CYP2D6 from single-point plasma, determine the antimodes between different phenotype groups, and validate the regression models obtained from small clinical trials.Results: Statistically significant correlations were found between MR from AUC and from serial plasma points from 1 to 30 h.The estimated density line with normal distribution reveals an antimode of 0.4 for log of the AUC metabolic ratio (log MRAUC) to distinguish EMs from IMs, and 2.0 to differentiate IMs from PMs.Both linear and power regression models were demonstrated to be acceptable as descriptions of the data.The linear regression model described the data better than the power model with larger R2 values.Conclusion: A single-point plasma sample from 1 to 30 h after single oral dose of 30 mg DM could be used for phenotyping of CYP2D6.PBPK models and large virtual populations are useful tools to evaluate phenotyping methods of CYP2D6, determine antimodes between different phenotype groups, and validate regression models.