The analysis of B-cell triggering mechanisms and clonal regulation has been successfully approached with ligands that are competent to activate a large fraction of all cells (1-3). Polyclonal activation of B cells has also proved to be a useful tool in studying Ig synthesis and secretion (4), in the analysis of functional B-cell subpopulations (5) and their development (6), and in the description of the antibody repertoire (7,8) by allowing the determination of relative frequencies of clonotypes in the absence of antigen selection.Each B-cell mitogen known to date activates only a fraction of all bone marrowderived lymphocytes (9, 10). Although some discrepancies in quantitative aspects of the problem still exist, it is now clear that partially distinct subsets of B cells respond to the various mitogens (5,11,12). In particular, much attention has been focused on the existence of B-cell subpopulations responding to either thymus-independent or thymus-dependent forms of the same antigenic determinants (13-16). As was pointed out before (10), this cell population that secretes antibodies subsequently to T cellmacrophage help has thus far escaped all analyses using direct B-cell mitogens. The study of this cell subset would provide new information on both its antibody repertoire and the mechanism of cooperative B-cell triggering which, to date, remains controversial (17).We attempted to develop a system for polyclonal B-cell induction mediated by T cells, in the absence of extraneously added ligands, such as antigen or mitogen. Previous experiments by Cammisuli et al. (18) had already indicated that cooperative B-cell induction could take place in the absence of specific interactions involving Ig receptors on the responding B cell, by using helper cell recognition of determinants artificially attached to B-cell surfaces. We have now improved that approach with three major modifications. To ensure that the antigens would remain on the B-cell surface throughout the cooperative culture, we activated T cells directly against minor antigens expressed on B-cell surfaces. Furthermore, we used an in vitro system by which a manifold enrichment of specific helper cells can be achieved. Finally, we measured the responses of all B cells, regardless of their antibody specificity. In this way, a very high frequency of specific T-B cell interactions was obtained in the helper assay, and this might make the system an optimal tool for the study of T-B cell collaboration.