We have recently demonstrated in the rat species that sponge matrix allograft infiltrating cells spontaneously produce nitric oxide (.N = 0) and this .N = 0 production precedes the development of CTL. Compared with our experience in the mouse species, the CTL activity recovered from rat sponge grafts is of shorter duration and less intense. Assessment of the spontaneous .N = 0 production by mouse allograft infiltrating cells reveals a more delayed time course of production, paralleling the recovery of CTL activity from the graft. The in vitro spontaneous .N = 0 production by mouse allograft-infiltrating cells was greater than the production by syngeneic graft-infiltrating cells on all days tested. Exposure of allogeneic but not syngeneic graft infiltrating cells to the sensitizing alloantigen in vitro resulted in enhanced .N = 0 synthesis. In contrast, LPS stimulated .N = 0 production by both syngeneic and allogeneic graft cells on all days postgrafting. Culture of day-14 allograft infiltrating cells with alloantigen in the absence of NG-monomethyl-L-arginine (NMA), the competitive inhibitor of .N = 0 synthesis, resulted in elevated supernatant NO2- levels and decreased 3H-TdR uptake and CTL activity compared with cultures carried out in the presence of NMA. The supernatant NO2- levels, as well as the CTL activity and 3H-TdR incorporation of the cultured cells, was dependent on the concentration of NMA present, and these effects could be reversed by excess L-arginine. Thus, the species difference in .N = 0 synthesis (rat > mouse), observed by others, is evident in the sponge allograft model and may explain why CTL activity recovered from rat allografts is of shorter duration and less intense than that from the mouse allografts.