Contact hypersensitivity (CS) to 2,4-dinitrofluorobenzene (DNFB) in mice is a form of delayed hypersensitivity that can be measured both in vivo (ear swelling) and in vitro (antigen-induced proliferation). Experiments were done to determine whether in vivo and in vitro manifestations of the DH reaction were mediated by different T cell subpopulations. DNFB-immune LN cells were separated on nylon wool columns into nonadherent (T-enriched) and adherent (B-enriched) populations and tested for their ability to transfer CS to normal recipients and to respond to antigen stimulation in vitro. Unseparated LN cells transferred CS and gave a strong in vitro proliferative response to the antigen 2,4-dinitrobenzene sulfonate (DNBS). T-enriched LN cells also transferred CS but did not respond in vitro to either PHA or DNBS, although the response to PHA, but not antigen, could be restored by addition of peritoneal macrophages. B-enriched LN cells did not transfer CS, although they gave a strong antigen response in vitro which was T dependent. Additional experiments using anti-Ia serum plus C showed that the T cells that transferred CS were Ia-, whereas the T cells that responded to antigen stimulation in vitro were Ia+. Collectively, these results indicate that distinct T cell subpopulations mediate the in vivo and in vitro counterparts of this DH reaction. The two populations are separable by virtue of one adhering to nylon wool and by differential sensitivity to lysis by anti-Ia serum plus C. These findings are discussed in terms of the role of lymphocyte subpopulations in CS and with regard to in vitro methods used to investigate cellular and genetic requirements for T cell activation.