Precise definition of the role of both CD4 and CD8 T-cell subsets from NOD mice in the adoptive transfer of diabetes has been complicated by the possibility that endogenous T-cells may be recruited. Two newly created NOD congenic stocks, NOD.NON-Thy-1^aand NOD/LtSz-scid, have been used as T-cell donors and recipients, respectively, to eliminate contributions from endogenous T-cells and thus to define the requirement for transferred T-cell subsets as a function of underlying diabetes development in the NOD donor. Total T-cells and T-cell subsets prepared from either prediabetic or diabetic NOD.NON-Thy-1^a donors were adoptively transferred into 6-wk-old NOD-scid/scid recipients that were monitored for diabetes development. Both flow cytometric and histological analysis of recipient spleen and pancreas after adoptive transfer showed lymphocytes of donor (Thy1.1⁺) origin exclusively. Total T-cell and enriched CD4⁺ T-cell preparations from both diabetic and young prediabetic donors transferred diabetes to NOD-scid/scid recipients. However, the mean time to diabetes onset was doubled when CD4⁺ lymphocytes were isolated from prediabetic versus diabetic donors, and t,⁺ cells over time. Enriched CD8⁺ populations alone were unable to transfer disease. More rigorous exclusion of CD8⁺ cells by means of anti-CD8 MoAb treatment in vivo of the recipients of enriched CD4⁺ cells demonstrated a significant difference in the diabetogenic potency of CD4⁺ lymphocytes from diabetic versus nondiabetic donors. Diabetes was adoptively transferred to 58% of the recipients of enriched CD4⁺ lymphocytes from diabetic donors. In contrast, none of the recipients of enriched CD4⁺ lymphocytes from young prediabetic donors developed diabetes after MoAb treatment in vivo. The ability of a T-cell population to produce severe insulitis and sialitis in NOD-scid/scid recipients of T-cells closely paralleled its ability to induce diabetes. In an effort to suppress insulitis by suppression of macrophage migration to the islets, NOD-scid/scid mice were treated with silica in conjunction with adoptive transfer of T-cells from diabetic donors. Chronic silica treatment failed to deplete tissue macrophages and did not prevent diabetes development after transfer of unfractionated T-cells. Evidence is discussed indicating that the age-associated differences in ability of CD4⁺ T-cells to adoptively transfer diabetes in the absence of the CD8⁺ T-cells subset is a function of prior, chronic exposure of the CD4⁺ lymphocytes to β-cell antigens in the donor. This study confirms that both CD4⁺ and CD8⁺ T-cells are required to initiate β-cell destruction in NOD mice.