Type 1 diabetes (T1D) is a prototypic organ-specific autoimmune disease resulting from the selective destruction of insulin-secreting β-cells within the pancreatic islets of Langerhans. It is caused by an immune-mediated inflammation, involving autoreactive CD4+ and CD8+ T lymphocytes that infiltrate the islets and initiate insulitis. The use of exogenous insulin is the current standard treatment. However, in spite of significant advances, this therapy is still associated with major constraints, including risk of hypoglycemia and severe degenerative complications. As T1D mainly affects children and young adults, any candidate immune therapy must be safe, and it must avoid a sustained depression of immune responses with all its attendant problems of recurrent infection and drug toxicity. In this context, inducing or restoring immune tolerance to target autoantigens would be the ideal approach. We refer to immune tolerance here as the selective damping of the damaging autoimmune response following a short treatment, while keeping intact the capacity of the host to respond normally to exogenous antigens. The therapeutic approach we discuss in this article originates from attempts to induce tolerance both to soluble antigens and tissue antigens (i.e. alloantigens and autoantigens) by using biological agents that selectively interfere with lymphocyte activation, namely polyclonal and monoclonal anti-T cell antibodies. The challenged dogma was that, in an adult-primed immune system, it was not possible to restore self-tolerance therapeutically without the use of exogenous autoantigen administration. The reality has been that, in diabetes, endogenous host autoantigen can fulfill this role because a significant amount of functioning β-cells remains, even at the time of established hyperglycemia. Experimental results obtained in the 1990s showed that a shortterm CD3 antibody treatment in recently diagnosed diabetic non-obese diabetic (NOD) mice induced permanent remission of the disease by restoring self-tolerance. Based on these findings, phase I, II, and III trials were conducted using two distinct humanized Fc-mutated antibodies to human CD3, namely ChAglyCD3 (otelixizumab) and OKT3γ1 Ala- Ala (teplizumab). Overall, when dosing was adequate, the results demonstrated that CD3 antibodies preserved β-cell function very efficiently, maintaining significantly high levels of endogenous insulin secretion in treated patients for up to 24 months after treatment. These data provided the first proof of concept for a long-term therapeutic effect in T1D following a short course administration of a therapeutic agent. Our aim is to review these data and to discuss them in the context of the pitfalls linked to pharmaceutical development, especially in the context of pediatric patients, as in autoimmune diabetes.