Purpose: This phase I infusion rate escalation trial was undertaken to evaluate the maximum applicable infusion rate for rituximab without steroid premedication in patients having received one previous rituximab infusion. Experimental Design: Cohorts of at least three patients were assigned to rituximab with or without concomitant chemotherapy. The initial infusion rate was 200 mg/h in the first cohort, and was increased by 100 mg/h in each subsequent cohort to a maximum of 700 mg/h. In each patient the infusion rate was increased by 100 mg/h every 30 minutes to the total dose (375 mg/m 2 ). In the first six cohorts (21 patients), two well-tolerated rituximab administrations were required; in the 7th cohort (11 patients) one previously well-tolerated rituximab infusion was required. Patients did not receive steroid premedication and were monitored with electrocardiograms (ECG), echocardiograms, Holter ECGs, troponin, and brain natriuretic peptide (BNP). Results: Thirty-two patients were included and 128 cycles were done, 85 at a rate of 700 mg/h. Patients tolerated infusion rates without major side effects. There were no new clinically relevant ECG alterations. Troponin (< 0.1ng/L) and mean cardiac ejection fraction (65%) remained in the reference range; BNP baseline level increased significantly 24 hours after rituximab administration (from 30.4 to 64.1ng/L; P < 0.0001).Conclusions: Rituximab can be administered safely at 700 mg/h without steroid premedication in patients having received at least one rituximab dose in the previous 3 months.Rituximab is a monoclonal antibody against the lymphocyte surface antigen CD20. First introduced in 1995 for the treatment of non -Hodgkin's B-lymphoma, rituximab has today a broad spectrum of indications, including treatment of rheumatoid arthritis, thrombotic thrombocytopenic purpura, and systemic lupus erythematosis, among many others (1 -6). Treatment with rituximab is generally well tolerated and can be combined with chemotherapy, improving response rate, response duration, and in some cases overall survival of patients with B-cell lymphomas (1). Infusion-related reactions can occur in approximately one fourth of patients during the first administration, probably due to a release of cytokines into the blood as a consequence of the rapid destruction of circulating B cells (7). This is why it is recommended to choose a low initial infusion rate. The incidence and the severity of infusion-related side effects consistently decrease in the successive administrations; indeed after the second rituximab infusion almost no B cells are left in the blood for at least 2 to 3 months. The reconstitution begins usually after 6 months and is completed after a median of 12 months (8, 9). We hypothesized that no infusion-related reaction should occur from the second infusion, and that a faster application of rituximab could be considered safe. Although the duration of the infusion is usually 4 to 5 hours for the first administration, the present study was undertaken to evaluate i...