Although both vitamin A (VA) deficiency and aging are independently associated with alterations in immune function, the effects of marginal VA status or VA supplementation on the immune system during aging were not studied. A long-term dietary study was conducted in a rat model of aging to quantify changes in T-cell populations in blood and spleen, including T-cells bearing a marker of natural killer (NKT) cells. The study included nine treatment groups [three levels of dietary VA: marginal (0.35 RE/kg diet), control (4.0 RE/kg diet), and supplemented (50 RE/kg diet); and three age groups: young (2-3 mo), middle-aged (8-10 mo), and old 20-22 mo); diets were fed continuously from weaning to the end of the study period. CD3(+)/CD4(+) T-cells decreased in percentage and number in blood with age, CD8(+) cells increased (%), and the CD4/CD8 ratio decreased. Conversely, aging was associated with increased NKT cells (phenotype CD3(intermediate)/NKR-P1(+)). Based on regression analysis of flow cytometry data, the phenotype of most NKT cells was CD3(intermediate)/NKR-P1(+)/CD28(-). NKT cells, which are most likely of extrathymic origin, accounted for most of the decrease in the CD4/CD8 ratio. Marginal VA status, particularly in older rats, was associated with increases in the percentage of CD8(+) T cells, percentage and number of NKT cells, and peripheral blood cell anti-CD3epsilon-stimulated proliferative response, and decreases in the CD4/CD8 T-cell ratio and splenic cell interleukin-2 production. These differences and the reciprocal changes observed for NKT cells vs. T- and classical NK cells in aging VA-marginal rats suggest that low VA status during aging may increase the risk of infectious or neoplastic diseases that require a normal balance of T-cell or NK-cell responses.