Background: Arsenic trioxide (As2O3) is an anticancer drug that has been reported to induce apoptosis and inhibit differentiation in human plasmacytoma and normal plasma/B cells without significant myelosuppression. We assessed the ability of As2O3 as single therapy or in combination with an anti-CD20 monoclonal antibody (mAb) and whole body irradiation (WBI) to deplete B and plasma cells, both in vitro and in vivo, and to reduce the level of anti-αGall -3Gal antibody (anti-Gal Ab) in baboons. Methods: In vitro the effect of As2O3 on antibody secretion (anti-Gal IgM, total IgG and IgM) was measured by enzyme-linked immunospot assay (ELISPOT). Its inhibition of proliferation of baboon splenocytes and the NCI-H929 human plasmacytoma cell line was measured by tritiated thymidine uptake. In vivo: all baboons (n = 7) had undergone splenectomy. The effects of As2O3 (0.18 to 0.36 mg/kg) on B/plasma cell depletion and anti-Gal Ab production were assessed in three baboons. For comparison, three baboons received either WBI (2 × 150 cGy) or anti-CD20 mAb (20 mg/kg × 4 doses), or both WBI and anti-CD20 mAb. A final baboon received As2O3 + WBI (150 cGy) + anti-CD20 mAb. Anti-Gal Ab levels were measured daily by ELISA. Depletion of B cells from blood and bone marrow (BM) was monitored by flow cytometry and by histology of lymph nodes (LN). Autopsy was performed in three baboons. Results: In vitro: As2O3 (at 5 × 10-6 mol/l) reduced anti-Gal IgM and total IgM secretors by 76% (P = 0.53) and 95% (P < 0.001), respectively, but did not reduce total IgG secretors. As2O3 inhibited in a dose-dependent manner the proliferation of activated splenocytes and of the NCI-H929 plasmacytoma cell line; complete inhibition was achieved at a dose of 1 × 10-5 mol/l. In vivo: As2O3 was found to be toxic at the doses given and was associated with the deaths of two of the four baboons that received it. Daily intravenous therapy with As2O3 alone reduced B cells (CD20+) in the blood (by 50 to 90%), BM (40%) and LN (20 to 30%), but anti-Gal Ab levels were not significantly decreased. Anti-CD20 mAb therapy alone or WBI alone depleted B cells by 100% in the blood and BM, and 80 to 100% in the LN. The combination of anti-CD20 mAb + WBI led to depletion of B cells in blood, BM and LN for 3 months, but reduction of anti-Gal Ab remained marginal. The combination of As2O3 + anti-CD20 mAb + WBI did not reduce anti-Gal Ab levels further. At autopsy in the latter baboon, B cells remained present in Peyer's patches and tonsils. Conclusions: In vitro: As2O3 reduced B/plasma cell numbers and suppressed IgM secretors, but not IgG secretors. In vivo: As2O3 was not as effective as either anti-CD20 mAb or WBI in depleting B/plasma cells, and was largely ineffective in reducing anti-Gal Ab levels. Its administration was associated with considerable toxicity. Autopsy in one baboon suggested that B cells in Peyer's patches and tonsils may be resistant to therapy and remain a source of continuing production of anti-Gal Ab.