Rituximab is a second‐line option in adults with immune thrombocytopenia (ITP), but the estimated 5‐year response rate, only based on pooled retrospective data, is about 20%, and no studies have focused on long‐term safety. We conducted a prospective multicenter registry of 248 adults with ITP treated with rituximab with 5 years of follow‐up to assess its long‐term safety and efficacy. The median follow‐up was 68.4 [53.7‐78.5] months. The incidence of severe infections was only 2/100 patient‐years. Profound hypogammaglobulinemia (<5 g/L) developed in five patients at 15 to 31 months after the last rituximab infusion. In total, 25 patients died at a median age of 80 [69.5‐83.9] years, corresponding to a mortality rate of 2.3/100 patient‐years. Only three deaths related to infection that occurred 12 to 14 months after rituximab infusions could be due in part to rituximab. At 60 months of follow‐up, 73 (29.4%) patients had a sustained response. On univariate and multivariate analysis, the only factor significantly associated with sustained response was a previous transient response to corticosteroids (P = .022). Overall, 24 patients with an initial response and then relapse received retreatment with rituximab, which gave a response in 92%, with a higher duration of response in 54%. As a result of its safety profile and its sustained response rate, rituximab remains an important option in the current therapeutic armamentarium for adult ITP. Retreatment could be an effective and safe option.