Osteoclasts are responsible for bone destruction in rheumatoid arthritis (RA), and adipose-derived mesenchymal stromal cells (ADSCs) can inhibit experimental collagen-induced arthritis model. This study aims to determine whether ADSCs also suppresses osteoclastogenesis and bone erosion in collagen-induced arthritis (CIA). Osteoclasts were induced from bone marrow-derived CD11b+ cells with receptor activator of nuclear factor-κ B ligand (RANKL) and macrophage colony-stimulating factor (M-CSF) stimulation and assessed with tartrate-resistant acid phosphatase (TRAP) staining. For human cells, osteoclasts were produced from human CD14+ cells. ADSCs were generated and added to cultures with different ratios with CD11b+ cells. Transwell and antibody blockade experiments were performed to define the mechanism of action. NF-κB and RANKL expression were determined by Western blotting and RT-qPCR. About 2 × 106 ADSCs or fibroblast cells were adoptively transferred to DBA1/J mice on day 14 after immunization with type II collagen/complete Freund's adjuvant (CII/CFA) while the onset and severity of the CIA were monitored. Adipose-derived mesenchymal stromal cells but not fibroblast cells completely suppressed osteoclastogenesis in vitro for human and mice. ADSCs injected after immunization and before of onset of CIA significantly suppressed disease development. Treatment with ADSCs dramatically decreased the levels of NF-κB p65/p50 in osteoclasts in vitro and P65/50 and RANKL expression by synovial tissues in vivo. We have demonstrated that ADSCs can inhibit RANKL-induced osteoclasts genesis via CD39 signals. Our findings also suggest that ADSCs can inhibit osteoclasts genesis without the involvement of regulatory T cells. ADSCs might represent a promising strategy for stem cell-based therapies for RA. Thus, manipulation of ADSCs may have therapeutic effects on RA and other bone erosion-related diseases.