Cytotoxic CD8+ T lymphoctyes (CTL) mainly utilize perforin and granzymes to kill beta cells, and perforin deficiency (pfp-/-) protects non-obese diabetic (NOD) mice from type 1 diabetes (T1D)1. In contrast, perforin deficiency in a CD8+ T cell-dependent model of diabetes, NOD8.3, had no effect on diabetes1 . This suggests that perforin may be required for CTL development and/or function in addition to its cytotoxic role, or that perforin produced by other immune cells might be required for T1D. Understanding the reason for this difference is important to determine if inhibiting perforin will be effective for T1D prevention or cure. NOD Rag1-/- mice that received wild-type CD4+ T cells complemented with pfp-/- CTL were significantly (p=0.005) protected from T1D compared to mice that received both wild-type CD4+ T cells and CTL, indicating that perforin deficiency in CTL is responsible for protection from T1D in NOD mice. However this result does not explain why NOD8.3pfp-/- mice had a similar incidence of T1D to NOD8.3 mice. We hypothesized that in NOD8.3pfp-/- mice, CTL-mediated elimination of APC, resulting in immune regulation, was lost. We observed a similar number of APCs in NOD8.3pfp-/- mice compared to NOD8.3 mice, and proliferation of CTLs in the islets and pancreatic lymph nodes was also similar. This indicates that perforin deficiency doesn't affect antigen presentation in NOD8.3 mice. Finally we hypothesized that perforin might be playing a previously unappreciated role in regulating naïve T cell responses. We transferred 8.3 T cells and 8.3pfp-/- T cells into NOD-IGRP mice that express IGRP in their APC. Perforin-deficient 8.3 T cells proliferated with accelerated kinetics compared to 8.3 T cells, suggesting that in the absence of perforin, 8.3 T cells are hyperactivated in response to antigen, and these CTL then use perforin-independent mechanisms to kill beta cells resulting in T1D.