IL-2 and TGF-β1 play key roles in the immunobiology of Foxp3-expressing CD25(+)CD4(+) T cells (Foxp3(+)Treg). Administration of these cytokines offers an appealing approach to manipulate the Foxp3(+)Treg pool and treat T cell-mediated autoimmunity such as type 1 diabetes. However, efficacy of cytokine treatment is dependent on the mode of application, and the potent pleiotropic effects of cytokines like IL-2 may lead to severe side effects. In the current study, we used a gene therapy-based approach to assess the efficacy of recombinant adeno-associated virus vectors expressing inducible IL-2 or TGF-β1 transgenes to suppress ongoing β cell autoimmunity in NOD mice. Intramuscular vaccination of recombinant adeno-associated virus to 10-wk-old NOD female mice and a subsequent 3 wk induction of IL-2 was sufficient to prevent diabetes and block the progression of insulitis. Protection correlated with an increased frequency of Foxp3(+)Treg in the periphery as well as in the draining pancreatic lymph nodes and islets. IL-2 induced a shift in the ratio favoring Foxp3(+)Treg versus IFN-γ-expressing T cells infiltrating the islets. Induction of IL-2 had no systemic effect on the frequency or activational status of T cells and NK cells. Induction of TGF-β1 had no effect on the Foxp3(+)Treg pool or the progression of β cell autoimmunity despite induced systemic levels of activated TGF-β1 that were comparable to IL-2. These results demonstrate that inducible IL-2 gene therapy is an effective and safe approach to manipulate Foxp3(+)Treg and suppress T cell-mediated autoimmunity and that under the conditions employed, IL-2 is more potent than TGF-β1.