Rituximab (RTX) efficacy in NMO is suggested by several case series. No consensus exists on optimal dosing strategies. At present the treatment schedules more frequently used are 375 mg/m2/week iv for 4 weeks (RTX-A) and 1000 mg iv twice, 2 weeks apart (RTX-B). Aim of this study is to confirm RTX efficacy and safety in the treatment of NMO and to evaluate whether a most favourable dosage regimen exists. Data on RTX-treated NMO patients were collected from 13 Italian Hospitals. 73 patients (64 F), were enlisted. RTX-A was administered in 42/73 patients, RTX-B in 31/73. Median follow-up was 27 months (range 7-106). Mean relapse rate in the previous year before RTX start was 2.2 ± 1.3 for RTX-A and 2.3 ± 1.2 for RTX-B. ARR in the first year of treatment was 0.8 ± 0.9 for RTX-A and 0.2 ± 0.4 for RTX-B, in the second year of treatment was 0.9 ± 1.5 for RTX-A and 0.4 ± 0.8 for RTX-B patients (p = 0.001 for the first year, ns (0.09) for the second year). RTX-B was more effective in delaying the occurrence of a relapse (HR 2.2 (95 % IC 1.08-4.53) p = 0.02). Adverse events were described in 19/73 patients (mainly urinary tract and respiratory infections, and infusion reactions). Two deaths were reported in severely disabled patients. Though with the limitations of an observational study, our data support RTX efficacy in NMO and suggest that high dose pulses might be more effective than a more fractioned dose.