We have developed a method which allows us to clone and reclone primed responder T cells derived from serially restimulated murine mixed lymphocyte cultures. We have derived clones from two such mixed lymphocyte cultures, A anti-B6 [A(B6)] and A anti-(B6XA)F1 [A(B6A)]. In the A (B6) system, we have isolated clones which can ve stimulated by B6 but not by (B6XA)F1 cells. This implies the presence of a unique parental H-2b MLR determinant which is absent on semi-allogeneic (B6XA)F1 cells. In the A(B6A) system, we have isolated clones which can be stimulated by (B6XA)F1 cells but not by B6 cells. This confirms our previous observation on the presence of unique hybrid MLR stimulating determinants on (B6XA)F1 cells. Many of the "clones" derived primarily from soft agar seem to be contaminated and contain several different sets of primed responder cells with different reactivity patterns . Experiments in which we subclone cells exhibiting selected reactivity patterns from such contaminated primary clones suggested that a T-cell growth factor or accessory cell is required for proliferation in soft agar following all antigen recognition by primed responder cells.