Type I diabetes (T1D) is caused by immune-mediated destruction of pancreatic beta cells. This process is triggered, in part, by specific (aa 9-23) epitopes of the insulin Β chain. Previously, fish insulins were used clinically in patients allergic to bovine or porcine insulin. Fish and human insulin differ by two amino acids in the critical immunogenic region (aa 9-23) of the B chain. We hypothesized that β cells synthesizing fish insulin would be less immunogenic in a mouse model of T1D. Transgenic NOD mice in which Greater Amberjack fish (Seriola dumerili) insulin was substituted for the insulin 2 gene were generated (mouse Ins1-/- mouse Ins2-/- fish Ins2+/+). In these mice, pancreatic islets remained free of autoimmune attack. To determine whether such reduction in immunogenicity is sufficient to protect β cells from autoimmunity upon transplantation, we transplanted fish Ins2 transgenic (expressing solely Seriola dumerili Ins2), NOD, or B16:A-dKO islets under the kidney capsules of 5 weeks old female NOD wildtype mice. The B:Y16A Β chain substitution has been previously shown to be protective of T1D in NOD mice. NOD mice receiving Seriola dumerili transgenic islet transplants showed a significant (p = 0.004) prolongation of their euglycemic period (by 6 weeks; up to 18 weeks of age) compared to un-manipulated female NOD (diabetes onset at 12 weeks of age) and those receiving B16:A-dKO islet transplants (diabetes onset at 12 weeks of age). These data support the concept that specific amino acid sequence modifications can reduce insulin immunogenicity. Additionally, our study shows that alteration of a single epitope is not sufficient to halt an ongoing autoimmune response. Which, and how many, T cell epitopes are required and suffice to perpetuate autoimmunity is currently unknown. Such studies may be useful to achieve host tolerance to β cells by inactivating key immunogenic epitopes of stem cell-derived β cells intended for transplantation.