We have used a T-cell receptor transgenic mouse model to study the role of antigen in the changes that occur as T cells age. We find that the characteristic shift in the CD4 population to a predominance of memory phenotype T cells which accompanies aging in non-transgenic mice does not occur, suggesting that this shift is a result of antigenic stimulation. Thus at least one component of aging must be antigen dependent. When responses of naive CD4 T cells from aged and young mice are directly compared in vitro, the former are relatively deficient in their ability to produce IL-2 and IL-3, they express altered levels of P-glycoprotein and they proliferate less well in the absence of exogenous cytokines. When the ability of both naive populations to generate effectors is compared, the number of effectors generated from aged naive cells is much reduced and the effectors generated express lower levels of IL-2R alpha and produce reduced levels of cytokines. Importantly, addition of IL-2 restores proliferation of aged naive T cells, restores efficient effector generation and results in effectors seemingly indistinguishable from those derived from young CD4 cells. Similar phenotypic and functional changes seen with aging are also found in T-cell populations from IL-2 and IL-2R alpha knockout mice. Thus the loss of optimal IL-2 production may participate in the aging process and may represent the main antigen-independent defect in the CD4 T-cell population.