Purpose: To accelerate early phase clinical development of a novel drug, teriflunomide sodium, to treat systemic lupus erythematosus (SLE) based on the data of leflunomide. Methods: Based on a pharmacokinetic (PK) study assessing the relative bioavailability of teriflunomide sodium compared to leflunomide, a population pharmacokinetic (Pop PK) analysis was firstly conducted using non-linear mixed effect model. Covariates were thoughtfully screened after Pop PK model evaluation and qualification using various diagnostic plots, visual predicted check (VPC) and bootstrap method. In order to predict teriflunomide PK profiles for multiple dosing of teriflunomide sodium in SLE patients, a model integrating enterohepatic circulation (EHC) mechanism was utilized to simulate the teriflunomide PK profile after multiple dosing of 20 mg/day leflunomide, and compare it to the teriflunomide PK profile in a 20 mg/day leflunomide multiple dose study in rheumatoid arthritis patients. Validated EHC PK model was applied to optimize dose regimen for teriflunomide sodium in SLE patients. Results: A population one-compartment model with pulsed EHC characteristic was developed to capture teriflunomide PK profiles after administration of leflunomide and teriflunomide sodium. Body weight and male sex were found to significantly increase apparent volume of central compartment. ABCG2 34G > A polymorphism was found to significantly change apparent clearance and absorption rate. The Pop PK model was evaluated and validated. After this model was confirmed to capture EHC characteristics of teriflunomide in both healthy subjects and patients with rheumatoid arthritis after single and multiple dosing leflunomide, it was applied to suggest dose regimen of teriflunomide sodium in phase II study. Conclusions: The pulsed EHC Pop PK model characterized the teriflunomide PK processes well in both healthy subjects and patients. Body weight, sex, and ABCG2 34G > A genotype were identified to significantly affect PK characteristics. The developed EHC Pop PK model exhibited the ability to predict PK profiles of teriflunomide in patients after long-term dosing and could be utilized to support phase II trial design.