Rivaroxaban is a popular direct factor Xa inhibitor used for anticoagulation therapy in patients with nonvalvular atrial fibrillation (NVAF). The aim of this study was to establish a population pharmacokinetic (PPK) model for rivaroxaban in elderly Chinese patients with nonvalvular atrial fibrillation, evaluate precision dosing regimens, and analyze hemorrhagic risk after rivaroxaban treatment. A 1-compartment population PK model with estimated glomerular filtration rate (eGFR), total bilirubin (TBIL), and ABCB1 rs1045642 as major covariates for apparent clearance was developed using the nonlinear mixed-effects model (NONMEM). A Monte Carlo simulation was performed to evaluate various dosing schemes and different levels of covariates for the target range of therapeutic drug-monitoring concentrations (C max,ss and C min,ss ). The exposure to rivaroxaban was simulated and assessed through hemorrhagic risk evaluation. The results showed that the average probability of target attainment (PTA) for optimal dosing regimens with different covariate levels for the targeted C max,ss and C min,ss were 29.35% to 31.3% and 64.91% to 65.8%, respectively. A dosage of 10 mg of rivaroxaban in elderly Chinese patients with normal renal and liver function was appropriate. The area under the concentration-time curve estimated over 24 hours with precision dosing at steady state (AUC 24,ss ) was statistically significantly associated with an increased risk of bleeding events (OR 1.0006, 95%CI 1.0003 to 1.001, P < .0001), and the bleeding risk increased by 1.82-fold for every 1000 μg*h/L increase in AUC 24,ss . A lower dose is recommended for elderly patients with renal impairment to avoid overexposure and bleeding events. The PPK model could inform individualized dosing for elderly Chinese patients with nonvalvular atrial fibrillation receiving rivaroxaban anticoagulation therapy.