B16 melanoma-bearing mice were treated with anti-interleukin 4 antibody, indomethacin or its combination to evaluate the ability of the primary tumor to induce lung metastasis and the antitumor host response. Flow cytometry of tumor cells incubated with sera from tumor-bearing mice showed B16 melanoma to induce a significant antitumor humoral response (39.0 +/- 1.1% positive cells versus 1.8 +/- 0.9% in the control). The treatment of tumor-bearing mice with antimouse anti-interleukin 4 monoclonal antibody plus indomethacin significantly increased (P < .01) the mean value of lung metastasis (from 6.1 +/- 3.0 in the controls to 50.8 +/- 21.8). Also, a significant increase in natural cytotoxicity against tumor cells was observed when both peripheral blood mononuclear cells and splenocytes were used as effector cells. In contrast, an antibody-dependent cellular cytotoxicity decrease was found with effector cells from both normal and tumor-bearing mice. In the former, the antibody-dependent cellular cytotoxicity decrease was 49.4% and 58.4% (P < .05) for peripheral blood mononuclear cells and splenocytes, whereas in the second case the decrease was 40.7% (P < .05) and 29.1% (P < .01), respectively. These results suggest that an efficient antibody-dependent cellular cytotoxicity response might be necessary to secure an effective host antitumor immune response.