Despite decades of research using various animal models for type 1 diabetes, we are still struggling to define the initiating autoantigens, the precise mechanisms of beta cell destruction, and suitable immune-based interventions to prevent or treat human diabetes. Animal models, such as the non-obese diabetic (NOD) mouse and the biobreeding (BB) rat, develop immune-mediated diseases with features resembling type 1 diabetes in humans. Although these animal models of autoimmune diabetes have proved to be valuable tools to study certain aspects of the disease process, they have also led to misconceptions and erroneous extrapolations, as well as false expectations with regard to the efficacy of immunotherapy. It is therefore time to ask ourselves whether we are making major strategic mistakes when employing rodent models for the study of type 1 diabetes. This review will describe where rodent models have provided us with proper guidance and where they have misled us, concluding that each model only offers partial information with undefined clinical value. Therefore, a more critical attitude and repetition of crucial observations in different model settings will be necessary in the future. I will argue that animal models have limited but evident value when it comes to teaching us about type 1 diabetes in humans, and we can take advantage of this value more efficiently.