Four biological response modifiers (BRMs), MVE-2 (maleic anhydride divinyl ether), Corynebacterium parvum (C. Parvum), PolyICLC (polyinosinic:polycytidylic acid stabilized with poly-L-lysine), and mouse alpha beta-interferon (alpha beta-IFN), were tested to assess whether repeated treatments would repeatedly induce or sustain augmented levels of natural killer (NK) cell activity and/or macrophage (M0)-mediated inhibition of tumor cell growth. In contrast to a significant increase in splenic NK activity obtained with a single treatment with each of the agents, multiple treatments with these BRMs led to a progressive decrease in the degree of augmentation of NK activity. In contrast, multiple injections with these agents resulted in sustained augmentation of M0-mediated reactivity. Separation of the spleen cells by Percoll discontinuous density gradient centrifugation indicated that with mice treated once with each BRM high levels of NK activity were detected in the lower density fractions and that these fractions contained a higher percentage of large granular lymphocytes (LGLs) than that found in comparable fractions from normal mice. In contrast, cells in the lower density fractions from mice that received multiple treatments had decreased NK activity and an appreciably lower proportion of LGLs. These results indicate that the development of hyporesponsiveness to augmentation of splenic NK-cell activity following multiple treatments with BRMs may be attributable to a decreased percentage of LGLs, the effector cell population responsible for NK cell-mediated cytotoxicity.