Insulin-dependent diabetes mellitus is associated with a growing number of immune abnormalities. At the time of clinical onset, most patients developing the disease as children or young adults have autoantibodies reactive with islet beta cells. Current autoantibody markers for IDDM are not sufficient to predict the disease in the general population. Studies in first-degree relatives indicate the presence of a subclinical disease characterized by beta cell dysfunction, which may or may not progress to overt IDDM. Although IDDM is genetically linked to certain HLA-DQ class II molecules, it needs to be clarified whether these molecules determine the propensity to react to certain antigens, the failure to maintain tolerance, or the ability to produce disease-associated autoantibodies. Circumstantial evidence suggests that yet another gene outside the HLA complex on chromosome 6 is more important. The interaction with the environment needs to be clarified, and the etiologic role of viruses has not been substantiated. An underlying systemic autoimmune propensity may influence environmental insults and perpetuate islet beta cell destruction. Until these mechanisms are understood, clinicians should periodically check their patients with IDDM for other organ-specific as well as non--organ-specific autoimmune diseases. Our understanding of these phenomena is poor, which may explain why clinical trials with immunosuppressive agents have been of limited success. Further studies on the molecular biology of the immune response against islet beta cell-specific antigens are necessary for the development of both predictive tests and novel measures to prevent IDDM.