Backgrounds: Rituximab 375 mg/m 2 weekly for 4 wks has significant activity in adults with primary immune thrombocytopenia (ITP). In this setting, several evidences support the possible use of lower doses of rituximab. Objectives: To investigate the activity of low-dose rituximab as salvage therapy in previously treated symptomatic ITP. Methods: Forty-eight adult patients were treated prospectively with rituximab 100 mg weekly for 4 wks. Results: Overall and complete responses (CR) (platelet level ≥ 50 and 100 x 10 9 /L) were 60.5% and 39.5%, respectively. In responders, the median time to response was 35 d (range: 7-112 d). The median time of observation was 18 months (range 3-49 months). Sixteen of 29 responding patients (55%) relapsed and 14 needed further treatments. The 12- and 24-month cumulative relapse-free survival was 61 % and 45%, respectively. In univariate analysis, CR rate was in inverse relation with weight OR = 0.95, CI 95 % [0.91; 0.99] (P = 0.019) and age OR=0.96, CI 95% [0.93; 0.99] (P = 0.047). Cox regression model showed that relapse probability increases as weight (HR = 1.06, CI 95 % [1.0031; 1.111]) and period between diagnosis and rituximab therapy (HR = 1.01, CI 95% , [1.002; 1.017]) increase. One patient developed an interstitial pneumonia 1 month after the end of rituximab treatment. No other infectious, hematologic or extra-hematologic complications were documented during follow-up. Conclusions: Low-dose rituximab is active in ITP but has moderate long-term effect. A comparative study with standard dose is warranted.