Objective: To study the clinical outcomes of low-dose leflunomide for rheumatoid arthritis (RA) complicated with hepatitis B virus (HBV) carriage and to observe the safety. Methods: A total of 115 RA patients were divided into three groups according to the state of HBV. They were all given leflunomide to observe the clinical outcomes and whether HBV was activated. Results: The indices (e.g. activity score) of all patients were significantly better after treatment than those before (P < 0.05), with 89.00% (92/115) of them reaching ACR20. Fourteen cases (12.2%) suffered from abnormal liver functions, and 5 cases who had HBV reactivation originated from the HBV carriage group. Neither the previous HBV infection group nor the infection-free group succumbed to HBV reactivation. The multiple regression model showed that the HBV reactivation risk of RA patients treated by leflunomide was increased by 30% by the basic state of hepatitis B as well as alanine transaminase level and swollen joint count before treatment. Conclusion: Leflunomide exerted satisfactory therapeutic effects on RA, but liver diseases, liver function, HBV-DNA load and the reactivation risks of carried HBV should be thoroughly checked and cautiously pondered.