In an attempt to resolve the issue of whether H-2-restricted T cell specificity is controlled by thymic epithelial cells or by cells of the macrophage/dendritic cell (M phi/DC) lineages, long-term F1----parent chimeras were subjected to secondary irradiation and reconstitution with F1 marrow cells. The rationale was that if F1 M phi/DC enter the thymus only quite slowly after irradiation, as claimed by other investigators, leaving F1----parent chimeras for a period of several months before re-irradiation would ensure that the new wave of T cells generated in the thymus of the chimeras would have no difficulty in making contact with donor-derived F1 M phi/DC. According to the view that M phi/DC rather than epithelial cells control H-2 restriction, the T cells differentiating in these chimeras would be expected to show H-2 restriction to both parental strains. In practice, T cells from twice-irradiated (1000 + 800 rad) chimeras showed strong restriction to host (thymic) H-2 determinants, the degree of restriction to host determinants being as marked as with T cells from once-irradiated chimeras. This finding applied both to T proliferative responses to KLH assayed in vitro and to T helper function for sheep erythrocytes measured in vivo. Preliminary experiments established that the initial dose of irradiation used for preparing the chimeras (1000 rad) resulted in almost total replacement of intrathymic M phi/DC by donor-derived cells within 4 wk of irradiation; M phi/DC were typed by determining their capacity to stimulate mixed-lymphocyte reactions. Collectively, the data imply that, at least under the conditions used, H-2-restricted T cell specificity is controlled by epithelial cells rather than by M phi/DC.