We have developed a new model to test the in vivo functions of NK cells. Injection of B6 mice with as little as 25 microliter of an NK-specific alloantiserum, anti-NK 1.1, significantly reduced the recipient's NK activity. The reduction, monitored in vitro as a decrease in the ability of spleen cells (SC) to lyse 51Cr-labelled YAC-1 target cells, occurred rapidly, within 2 h of administration of the anti-NK 1.1 serum. NK activity gradually returned to control levels but still was significantly depressed at 48 h. Comparable decreases were observed whether the serum was injected by the intravenous (i.v.) or the intraperitoneal (i.p.) route. Injection of the mice with exogenous complement (newborn rabbit serum) did not significantly increase the antiserum's effect. To test the requirement for NK cells in tumor clearance in vivo, mice were pre-treated with anti-NK 1.1 serum and subsequently injected i.v. with 125IdUrd (5-iodo 2'deoxyuridine)-labelled YAC-1 or RBL-5 lymphoma cells. Tumor cell clearance in lungs, liver and spleen was reduced two to four-fold in the anti-serum-treated mice compared to injection controls. These results provide direct evidence that NK cells are involved in the elimination of tumor cells in vivo.