It has been generally accepted that T cells play a critical role in endocrine autoimmune diseases. Immunotherapy aimed at T cells usually intervenes in the disease process. Yet, it has proven very difficult to identify the pathogenic T cells. This is partly caused by lack of measures to detect autoreactive T cells in a specific, sensitive and reproducible fashion as is achievable for determination of autoantibodies. There are, however, more explanations for the perhaps disappointing progress in this area: unlike autoantibodies, the relevant (disease-associated) autoreactive T cells act in the tissue lesion, and only circulate in very low precursor frequencies. Moreover, T cell autoreactivity is counteracted by various mechanisms of immune regulation. Candidate target autoantigens of T cells have been identified by autoantibodies, while there is little evidence that these autoantibodies are pathogenic. It is therefore conceivable that additional T cell targets exist. Finally, results from experimental animal models of endocrine autoimmunity have raised false expectations. Nonetheless, significant progress in our understanding of the contribution of (autoreactive) T cells to organ-specific destruction and autoimmune disease has been achieved. Although cross-sectional detection of autoreactive T cells bears little relevance to diagnosis, longitudinal studies have proven useful in determining the fate of islet implantation in type 1 diabetes patients, defined the immunological efficacy of immuno-intervention studies, and have led to definition of relevant target autoantigens and peptides that will help to monitor disease-associated autoimmunity. In conclusion, progress in the area of autoreactive T cells in the pathogenesis of type 1 diabetes may have seemed slow in the eyes of the beholder, but in fact, studies on T cells have contributed significantly to the unravelling of the pathogenic processes leading to the definition of appropriate targets for immuno-intervention.