The existence of T cells specific for soluble antigens in association with unique F(1) or recombinant major histocompatibility complex (MHC) gene products was first postulated from studies on the proliferative response of whole T cell populations to the antigen poly(Glu(55)Lys(36)Phe(9))(n) (GLφ). In this paper we use the newly developed technology of T lymphocyte cloning to establish unequivocally the existence of such cells specific for GLφ and to generalize their existence by showing that F(1)- specific cells can be isolated from T cell populations primed to poly(Glu(60)Ala(30)Tyr(10))(n) (GAT) where such clones represent only a minor subpopulation of cells. Gl.4b-primed B10.A(5R) and GAT-primed (B10.A × B10)F(1) lymph node T cells were cloned in soft agar, and the colonies that developed were picked and expanded in liquid culture. The GLφ-specific T cells were then recloned under conditions of high-plating efficiency to ensure that the final colonies originated from single cells. T cells from such rigorously cloned populations responded to stimulation with GILφ but only in the presence of nonimmune, irradiated spleen cells bearing (B10.A × B10)F(1) or the syngeneic B 10.A(5R) recombinant MHC haplotype. Spleen cells from either the B10 or B 10.A parental strains failed to support a proliferative response, even when added together. (B10 × B10.D2)F(1) and (B10 × B10.RIII)F(1) spleen cells also supported a proliferative response but (B10 × B10.Q)F(1) and (B10 X B10.S)F(1) spleen cells did not. These results suggested that the T cell clones were specific for GL[phi} in association with the β(AE)(b)-α(E) (k,d,r,) Ia molecule and that recognition required both gene products to be expressed in the same antigen-presenting cells. Support for this interpretation was obtained from inhibition experiments using the monoclonal antibody Y-17 specific for a determinant on the β(AE)(b)-αE Ia molecule. Y-17 completely inhibited the proliferative response of a GLφ-specific clone but had no effect on the response of either a PPD-specific or GAT-specific clone, both of which required the β(A)-α(A) Ia molecule as their restriction element. No evidence could be found for the involvement of suppressor T cells in this inhibition. We therefore conclude that the phenomenon of F(1)-restricted recognition by proliferating T cells results from the presence of antigen- specific clones that must recognize unique F(1) or recombinant Ia molecules on the surface of antigen-presenting cells in addition to antigen in order to be stimulated.