Abstract Background and Purpose: The 2019 novel coronavirus (COVID-19) has been spread out since December 2019 from China to 29 countries. No effective treatment is currently available, although the combination regimen of the antiretroviral drugs– lopinavir/ritonavir (LPV/r), with other antiviral drugs have been using, but the evidences are limited. A recent in vitro study showed that chloroquine could inhibit COVID-19 to cells, and enhance antiviral efficacy. This study aimed to predict the optimal dose regimens of LPV/r, and chloroquine in combination as a potential treatment of COVID-19 infection, using the physiologically-based pharmacokinetic (PBPK) modelling. Experimental approach: The whole PBPK models were constructed. The predicted plasma drug concentrations were compared with the published clinical data. The validated models were used to predict optimal dosage regimens of LPV/r, and chloroquine co-administration. The optimal dose regimen was determined based on the efficacy, and toxicity reported in the published data. Key Results: The average errors of the predicted values were within 30% of the observed data. The proposed optimal dosage regimen is the once-daily dose of 800/200 mg LPV/r co-administered with chloroquine at a loading dose of 1,000 mg, followed by twice-daily dose of 500 mg for 8 doses on the second day, and the twice-daily dose of 400 mg for 18 doses. Conclusion and Implications: PBPK modelling successfully predicted pharmacokinetic profiles within an acceptable range of errors. The study provides a focus for clinical studies to confirm the efficacy of the proposed dosage regimen as a novel treatment for COVID-19 infection.