Background. Previous studies have shown that skin allograft survival can be augmented by the administration of donor spleen or donor bone marrow in antithymocyte serum (ATS) treated recipients. Because natural killer cells (NK) have been reported to possess immunoregulatory properties, we investigated whether the ability of donor spleen or bone marrow cells to enhance allograft survival was dependent on the presence of donor NK cells. Methods. Recipient (C57BL/6 × A/J)F1 strain mice (H2 haplotypes Kb/k, Ab/k, E-/k, Db/d)were treated with ATS on days -1 and +2 relative transplantation of a C3H (H-2k) skin allograft. On day +7, each recipient was randomly assigned to one of the following groups that received i.v. donor C3H cell infusions via the tail vein: 1) 5.0 × 107wild-type donor spleen cells (SPC; 2) 5.0×107spleen cells from C3H/HeJ-Lystbg-2J/+ mice (commonly called beige mice and have selectively impaired NK cell function); 3) 2.5×107wild-type donor bone marrow cells (BMC); 4) 2.5×107beige C3H bone marrow cells; and 5) no donor cell infusion (ATS controls). In another experiment, each recipient was randomly assigned to one of the following groups that received injections of: 1) 4.75×107spleen cells depleted of NK cells; 2) 2.5×106purified splenic NK cells; 3) a coinfusion of 5.0×107beige spleen cells and 2.5×106purified wild-type splenic NK cells. Results. Recipients infused with wild-type SPC exhibited significant augmentation of allograft survival compared with ATS controls. However, graft survival was reduced in recipients that were infused with spleen cells from beige mice compared with recipients infused with wild-type SPC (median survival time (MST): 38 vs. 92 days, P=0.02). In contrast, infusions of beige BMC augmented allograft survival as well as wild-type BMC (MST: 47 vs. 49 days, P=0.76). Furthermore, the ability of wild-type SPC to augment allograft survival was abrogated by the depletion of NK cells (MST=92 vs. 34 days, respectively, P=0.005). The coinfusion of beige SPC and purified splenic NK cells enhanced allograft survival as well as wild-type SPC (MST=56 days, P=0.65). Finally, recipients infused with purified NK cells did not experience increased graft survival compared to recipients that received no infusion (MST=29 vs. 33 days, respectively, P=0.6). Conclusions. Donor splenic NK cells are necessary, but not sufficient, for the extension of graft survival by infusion of donor splenocytes, suggesting that they may work in concert with another cell-type. In contrast, the extension of graft survival by donor bone marrow does not depend on the presence of donor NK cells.