Résumé : Objectives: To assess efficacy and safety of rituximab (RTX) as induction therapy, maintenance of remission and treatment of relapses in a cohort of IgG4-related disease (IgG4-RD) patients. Methods: Nationwide retrospective multicenter study of IgG4-RD patients treated with at least one course of RTX. Clinical, biological and radiological response, relapse rate and drug tolerance were analyzed. Kaplan-Meier curves were plotted and risk factors for relapse studied with a Cox regression model. Results: Among 156 IgG4-RD patients included in the French database, 33 received rituximab. Clinical response was noted in 29/31 (93.5%) symptomatic patients. Glucocorticoids withdrawal was achieved in 17 (51.5%) patients. During a mean follow-up of 24.8 ±21 months, 13/31 (41.9%) responder patients relapsed after a mean delay of 19 ±11 months after RTX. Active disease, as defined by an IgG4-RD Responder Index >9 before RTX, was significantly associated with relapse (HR = 3.68, 95% CI: 1.1, 12.6) (P = 0.04), whereas maintenance therapy with systematic (i.e. before occurrence of a relapse) RTX retreatment was associated with longer relapse-free survival (41 versus 21 months; P = 0.02). Eight severe infections occurred in 4 patients during follow-up (severe infections rate of 12.1/100 patient-years) and hypogammaglobulinemia 5 g/l in 3 patients. Conclusion: RTX is effective for both induction therapy and treatment of relapses in IgG4-RD, but relapses are frequent after B-cell reconstitution. Maintenance therapy with systematic RTX infusions is associated with longer relapse-free survival and might represent a novel treatment strategy. Yet, the high rate of infections and the temporary effect of RTX might be hindrances to such strategy.