The capacity of interferon (IFN) given once or repeatedly to augment natural killer (NK) cell activity in the spleen was tested in young and aged C57BL/6 mice. Multiple injections of 10(4) u IFN i.p. into 3-month-old or 12-month-old mice did not augment cytotoxicity mediated by NK cell activity; rather it reduced this activity in relation to the effect obtained with a single injection. When T and B cells or macrophages were removed in vitro, no restoration of NK-cell activity, and therefore no suppressor cells, could be found in this system. In a long-term experiment, B6 mice were chronically injected with 10(4) u of IFN or mock-IFN from 5 months of age until death. No difference in the modification of NK-cell activity was observed after IFN and mock-IFN, regardless of the duration of treatment. However, the capacity of spleen cells to generate cytotoxic T cells after allogeneic stimulation in vitro or in vivo was increased in IFN-treated mice. The median survival time was improved by 7 weeks through treatment with IFN compared to mock-IFN, but was reduced by 9 weeks compared to the survival after treatment with saline. No significant difference in the incidence of spontaneous lymphoma resulted from IFN treatment, but the percentage of tumors observed in mice receiving mock-IFN (68.1%) was higher than that in mice receiving saline (44.1%). These results indicate that, while a stimulating effect resulting in a delay in mortality and in tumor development can be attributed to IFN itself, the contaminating antigenic products are plausibly responsible for the deleterious effect of the total preparation.