Glutamic acid decarboxylase (GAD) is considered to be one of the strongest candidate autoantigens involved in triggering beta-cell-specific autoimmunity. The majority of recent onset type 1 diabetes patients and pre-diabetic subjects have anti-GAD antibodies in their sera, as do nonobese diabetic (NOD) mice, one of the best animal models for human type I diabetes. Immunization of young NOD mice with GAD results in the prevention or delay of the disease as a result of tolerizing autoreactive T cells. Autoimmune diabetes can also be prevented by the suppression of GAD expression in antisense GAD transgenic mice backcrossed with NOD mice for seven generations. These results support the hypothesis that GAD plays an important role in the development of T-cell-mediated autoimmune diabetes. However, there is some controversy regarding the role of GAD in the pathogenesis of diabetes. Whether GAD truly plays a key role in the initiation of this disease remains to be determined. The examination of the development of insulitis and diabetes in beta-cell-specific GAD knockout NOD mice will answer this remaining question.