The ability of immune cells to home to peripheral lymphoid organs or target tissues is critical to autoimmune patho-genesis, but it also makes these cells a promising vehicle for delivering therapeutic agents to the relevant tissues. Two reports in this issue apply this principle to suppress collagen-induced arthritis (CIA), a mouse model of rheumatoid arthritis. Nakajima and colleagues previously used adoptive transfer of transgenic T cells to block autoimmune responses in the CNS, and here they employ a similar strategy, transducing collagen-specific helper T cells with an antagonist of the cytokine IL-12. They introduce the modified T cells into mice that would otherwise develop CIA and show that the cells home efficiently to the synovium and block inflammation in situ. In contrast, their CNS-directed T cells, which express the same immunosuppressive transgene and are effective against a different autoim-mune disease, do not persist in the synovium and fail to suppress arthritis. In a complementary effort, Morita et al. show that dendritic cells (DCs), which control the proliferation and immunological properties of T cells, can also be used to deliver an appropriate regulatory protein. These authors generated DCs, transduced them with IL-4, and rein-troduced them into mice. The modified DCs can interact with collagen-specific T cells and drive them toward the Th2 phenotype, which is associated with tolerance to the corresponding antigens. Here again, tissue targeting is cru-cial, since the DCs must reach the spleen if they are to have their desired effect on T-cell development. Morita et al. show that delivery by intraperitoneal injection leads to the most efficient targeting of DCs to the spleen and confers the greatest protection from arthritis.