Environmental factors are the main contributors to type 1 diabetes (T1D) pathogenesis, yet they remain unidentified. Enteroviruses are proposed candidate triggers due to temporal correlations between infection and T1D autoimmunity and to detection of viral proteins in diseased islets. However, such correlations are not universal and may be relatively uncommon. Furthermore, evidence of a cause–effect relationship is lacking, as infection of non-obese diabetic mice with Coxsackie enteroviruses can either trigger or blunt disease. The proposed mechanisms are either non-antigen-specific (i.e. β-cell destruction and release of sequestered antigens, islet inflammation) or antigen-specific (i.e. epitope mimicry, by which immune responses to enteroviruses may be diverted against homologous β-cell antigens). The case for the latter mechanisms is even less stringent, as there is little evidence of promiscuous antigen recognition at the single T-cell level. Other infectious agents may thus be implicated. Demonstration of their role will require fulfilling the Koch's postulates, namely isolation of the agent preferentially in T1D patients, including before disease onset; and T1D induction when the agent is inoculated into mice. The same is needed for cross-reactive T cells to support epitope mimicry mechanisms. Generation of alternative (humanized) mouse models that could be challenged with candidate microbes is needed.