Cholera toxin (CT) and its B subunit (CT-B) are potent oral immunogens in vivo, although both strongly inhibit polyclonal lymphocyte activation in vitro. In order to help understand this paradox, we have studied the activation and proliferation of CT-specific T cells in vitro, by using CT-B-primed lymph node T cells as responders, concanavalin A-stimulated peritoneal macrophages as antigen-presenting cells (APCs), and various forms of CT-B as antigen. The results indicate that in many ways CT-specific T cells respond in a manner similar to that of T cells specific for other protein antigens: the degree of proliferation was proportional to the dose of antigen and APCs in the cultures, was antigen specific, and was H-2 restricted. APCs from genetic high-responder strains to CT stimulated significantly more proliferation in F1 (high x low) responder T cells than did APCs from low responder strains. However, there was a marked difference in the activation of CT-specific T cells when different forms of CT-B were used. Native CT-B stimulated little or no T-cell proliferation, wherease denatured CT-B or CT-B blocked by its ligand, GM1 ganglioside, stimulated T cells well. Addition of native CT-B to cocultures of primed T cells, APCs, and these latter stimulatory forms of CT-B inhibited the specific proliferative response to CT-B to varying degrees, depending on the ratio of the two forms in culture. We conclude that the ability of CT-B to inhibit T cells extends even to T cells specific for CT itself. Because of these inhibitory properties, processing of CT to nonbinding molecular forms or fragments must be an important prerequisite for the immune response to CT to occur in vivo, and such processing is likely to be important in the immune response to a variety of other enterotoxins as well.