Objective: Rheumatoid arthritis (RA) is a major autoimmune disease that causes synovitis and joint damage. Although clinical trials have been performed using interleukin-10 (IL-10), an antiinflammatory cytokine, as a potential treatment of RA, the therapeutic effects of IL-10 have been limited, potentially due to insufficient residence in lymphoid organs, where antigen recognition primarily occurs. This study was undertaken to engineer an IL-10-serum albumin (SA) fusion protein and evaluate its effects in 2 murine models of RA. Methods: SA-fused IL-10 (SA-IL-10) was recombinantly expressed. Mice with collagen antibody-induced arthritis (n = 4-7 per group) or collagen-induced arthritis (n = 9-15 per group) were injected intravenously with wild-type IL-10 or SA-IL-10, and the retention of SA-IL-10 in the lymph nodes (LNs), immune cell composition in the paws, and therapeutic effect of SA-IL-10 on mice with arthritis were assessed. Results: SA fusion to IL-10 led to enhanced accumulation in the mouse LNs compared with unmodified IL-10. Intravenous SA-IL-10 treatment restored immune cell composition in the paws to a normal status, elevated the frequency of suppressive alternatively activated macrophages, reduced IL-17A levels in the paw-draining LN, and protected joint morphology. Intravenous SA-IL-10 treatment showed similar efficacy as treatment with an anti-tumor necrosis factor antibody. SA-IL-10 was equally effective when administered intravenously, locally, or subcutaneously, which is a benefit for clinical translation of this molecule. Conclusion: SA fusion to IL-10 is a simple but effective engineering strategy for RA therapy and has potential for clinical translation.