As an attempt to experimentally induce antigen (Ag)-specific immunosuppression, we have previously created CD95 ligand (CD95L)-transduced dendritic cells (DC), which delivered apoptotic, but not activation, signals to CD4+ T cells in vitro in an Ag-dependent manner. We have also demonstrated that CD95L-transduced DC (termed killer DC) injected into syngeneic animals suppressed delayed-type hypersensitivity responses to an administered Ag. Based on these findings, we tested whether the injection of killer DC derived from A/J mice (H-2a) into allogeneic BALB/c recipients (H-2d) could prolong the survival of A/J-derived skin grafts by depleting A/J-reactive effector T cells. This attempt has not been successful. In this study, we elucidate the reasons for this failure, especially in terms of in vitro effects of killer DC on in vivo primed alloreactive T cells. We show that killer DC (i) failed to induce the proliferation of naive alloreactive T cells in a CD95/CD95L-dependent fashion, (ii) inhibited the proliferation of in vivo primed alloreactive T cells, (iii) killed relatively small fractions (up to 30%) of these T cells in vitro in a CD95/CD95L-dependent fashion and (iv) significantly, but incompletely, inhibited the generation of cytotoxic T-lymphocyte activities against A/J determinants. Thus, killer DC have significant, but modest, capacities to suppress in vitro alloimmune responses, which may not be sufficient to prolong the survival of alloskin grafts in a stringent allograft model. This study suggests that the current format of killer DC technology requires more modifications for its clinical application to prevent graft rejection.