In the humoral immune response to most antigens studied so far, the presence of helper T cells is required for the generation of antibody-forming cells from precursor B cells. In terms of the classical carrier-hapten systems described by Mitchison (1), B cells specific for a hapten can only become antibody-forming cells if helper T cells react with antigenic determinants on the carrier molecule that the hapten is linked to. This finding has later on led to the hypothesis that the close proximity brought about by an antigen "bridge" between B cell and T cell is the prerequisite for T-B cooperation (2). Under this assumption, several concepts have been developed to explain the nature of T-cell help itself. Among these are: the focussing of antigen by the T cell onto the B cell (2, 3), a membrane interaction between the two cells (4), and the production of mediators, antigen specific (5-7) or nonspecific (4,(7)(8)(9)(10)(11)(12)(13)(14), by the antigen-stimulated T cell. In vitro studies have shown that T-cell-replacing factor (TRF),I a nonantigen-specific mediator produced by activated T cells (15), can replace T cells in primary and secondary immune responses to heterologous erythrocytes (15, 16). The action of TRF and similar factors described by other authors (4, 7-14) has so far been demonstrated by restoring the immune response to heterologous erythrocytes in T-cell-deprived cultures by addition of preparations containing the factor. On the other hand, the experiments reported by Hartmann (17) have shown that T cells activated to erythrocyte antigens in vivo do, upon antigenic challenge in vitro, not only restore the response of B-cell cultures to the epitopes of the same erythrocytes, but also to those of erythrocytes from another species added simultaneously to the same cultures. Although this co-stimulation is best explained by the in situ generation and action of TRF, the function, and indeed the presence of this mediator in complete spleen cell cultures during the immune response remains to be demonstrated.In order to study this question under the most stringent conditions, we chose the anamnestic IgG reaction to soluble carrier-hapten conjugates in vitro. The pronounced T-cell dependence of the IgG response to such a conjugate and the possibility to stimulate T and B cells separately with carrier and hapten determinants make it possible to investigate the action of TRF in the immune response to soluble protein antigens.