New strategies are required to clinically exploit the ability of monoclonal antibodies to target tumor for lysis by cellular effector mechanisms. In this report we examine the therapeutic effects of 2B1, a bispecific monoclonal antibody with specificity for the extracellular domain of the c-erbB-2 oncogene product and the human Fc gamma receptor, Fc gamma RIII (CD16), describe the characteristics and limitations of this model, and examine the mechanisms underlying the observed responses. The model uses SK-OV-3 human ovarian carcinoma xenografts in scid mice. These cells are susceptible to 2B1-directed lysis by human peripheral blood lymphocytes or lymphokine-activated killer cells, and maintain c-erbB-2 expression in vivo. 125I-labeled 2B1 selectively accumulates in tumor, with a peak of 10.5% injected dose/g of tumor 24 h following its i.v. injection. However, the selectivity of this binding is lessened by 2B1 accumulation in the lungs and other normal organs and persistence in the blood. This is caused by antibody binding to murine lung, colon, stomach, and skin expressing the epitope recognized by the anti-c-erbB-2 component of 2B1 in tumor-bearing, but not normal mice. In treatment studies using various permutations of antibody, human peripheral blood lymphocytes or lymphokine-activated killer cells and interleukin 2, cellular therapy alone had minimal effects on SK-OV-3 xenograft growth, but significantly improved when 2B1 treatment was incorporated. Median survivals increased from 80 +/- 3.5 days with no therapy to 131 +/- 7.3 days following therapy with 100 micrograms 2B1, interleukin 2, and human peripheral blood lymphocytes, with 70% of animals exhibiting no evidence of tumor at day 150. These effects were preserved when the cells were administered in human serum. In contrast, human serum abolished the antitumor effects of 520C9, which is the parent anti-c-erbB-2 antibody of 2B1. Thus 2B1-based therapy has therapeutic effects, without obvious toxicity, despite the targeting of this antibody to normal murine tissues. Since combinations of 2B1 and interleukin 2 may have antitumor properties, mechanisms other than bispecific monoclonal antibody-promoted conjugation of c-erbB-2 antigen-expressing tumor to CD16-expressing effector cells may be involved.