Background: Lupus animal model has shown that arsenic trioxide (ATO), a treatment of acute promyelocytic leukaemia, could be effective in SLE. This is the first clinical study to determine the safety and efficacy of a short course of intravenous ATO in patients with active SLE. Methods: This phase IIa, open-label, dose-escalating study enrolled 11 adult SLE patients with a non-organ threatening disease, clinically active despite conventional therapy. Patients received 10 IV infusions of ATO within 24 days. The first group received 0.10 mg/kg per injection, with dose-escalating to 0.15 mg/kg in a second group, and to 0.20 mg/kg in a third group. The primary endpoint was the occurrence of adverse events (AEs) and secondary endpoints were the number of SLE Responder Index 4 (SRI-4) responders at week 24 and reduction of corticosteroid dosage. In an exploratory analysis, we collected long-term data for safety and attainment of lupus low disease activity state (LLDAS). Results: Four serious AEs occurred (grade 3 neutropenia, osteitis, neuropathy), 2 of which were attributable to ATO (neutropenia in the 2 patients treated with mycophenolate). Two patients suffered a severe flare during the last 4 weeks of the trial. At W24, five patients among 10 were SRI-4 responders. Overall, mean corticosteroid dosage decreased from 11.25 mg/day at baseline to 6 mg/day at W24 (P < 0.01). In the long term, 6 patients attained LLDAS at W52, which continued at last follow-up (median LLDAS duration 3 years, range 2-4). Conclusions: A short course of ATO has an acceptable safety profile in SLE patients and encouraging efficacy.