Background: Since the discovery of IA-2 as a major autoantigen in type 1 diabetes, the question arose as to whether other PTPs (protein tyrosine phosphatases) could act as diabetic autoantigens as well. A novel PTP, designated IA-2 B (phogrin; phosphatase homologue in granules of insulinoma) was isolated that has a high sequence similarity to IA-2. Since some studies suggested that auto- immunity to phogrin, rather than IA-2 may be more closely associated with the development of type 1 diabetes, we measured the frequency of anti-phogrin antibody in Korean patients with type 1 diabetes and compared it with that of anti-IA-2 antibody/ anti-GAD antibody. Methods: The anti-phogrin antibody and the anti-IA-2 antibody were measured by radioligand binding assays using in vitro transcribed and translated S-labeled phogrin and IA-2, respectively. Anti-GAD antibody was measured using a commercial radioimmunoassay kit (RSR, Cardiff, U.K.). The subjects in this study consisted of 41 patients with classical type 1 diabetes, 22 with slowly progressive type 1 diabetes, and 39 with type 2 diabetes. Their average mean age was 16.9 years, 37.9 years and 45.3 years respectively. Results: The prevalence of anti-phogrin antibody, anti-IA-2 antibody and anti-GAD antibody in classical type 1 diabetes was 24.4%, 26.8% and 51.2% respectively. That, in slowly progressive type 1 diabetes was 0%, 9.1% and 40.9% respectively. When the anti-GAD antibody assay and the anti-IA-2 antibody assay were combined, the prevalence of autoantibodies was 58.5% in classical type 1 diabetes and 50% in slowly progressive type I diabetes. However, the addition of the anti-phogrin antibody to the anti-GAD antibody/anti-IA-2 antibody measurement did not significantly increase the prevalence of autoantibody. The level of the anti- phogrin antibody was positively correlated with that of the anti-IA-2 antibody. The presence of the anti-phogrin antibody and the anti-IA-2 antibody was negatively correlated with the age at diagnosis. One patient with type 1 diabetes had the anti-phogrin antibody without the anti-IA-2 antibody. Conclusion: Combined measurement of the anti-phogrin antibody with the anti-IA-2 antibody/ anti-GAD antibody did not significantly increase the prevalence of autoantibodies in Korean patients with type 1 diabetes. In the majority of Korean type 1 diabetes patients, the anti-phogrin antibody appears to share epitopes with the anti-IA-2 antibody. However, a small proportion of type 1 diabetes patients may have a specific autoimmune response to phogrin.