Diabetes mellitus is comprised primarily of two clinically separate diseases: type 1 (T1D) and type 2 diabetes (T2D). T1D is a cell-mediated autoimmune disease directed against the beta cells and characterized by autoantibody (Ab) and T cell reactivity to islet proteins whereas, T2D is non-autoimmune. Despite the fact that the pathological process in autoimmune diabetes involves T cells, immune markers of diabetes have primarily centered on the presence of circulating serum islet autoantibodies. In two masked NIH sponsored workshops, our cellular immunoblotting T cell assay, which uses isolated human islets separated into 18 molecular weight fractions, has been validated to be able to distinguish T1D patients from controls with excellent specificity and sensitivity. In this study, we utilized the first workshop to select eight molecular weight fractions of human islets that were the most discriminatory between T1D patients and controls. Using these eight molecular weight fractions identified in the first workshop, we validated the preferential recognition of these 8 blot sections in a second workshop. We then re-calculated the sensitivity and specificity of the cellular immunoblotting assay for both workshops using only the data from these 8 blot sections.We observed increases in both sensitivity and specificity compared to the original workshop data for both workshops. The use of 8 instead of 18 molecular weight regions allows for a significant reduction in the amount of blood needed from patients, thus allowing cellular immunoblotting to be performed on pediatric patients participating in immunomodulatory studies. This improved T cell assay, which directly measures islet reactive T cell responses in autoimmune diabetes patients with excellent sensitivity and specificity, will likely improve patient follow-up during intervention studies.