Monoclonal antibodies of murine origin have been administered to large numbers of patients with cancer. In general, these antibodies have been able to localize to sites of tumor but have not brought about meaningful tumor responses. The ultimate use of such antibodies will be limited by the development of human antimouse antibody responses. Such antibodies may not only cause some toxic effects but have been shown to influence the biodistribution of antibodies and the ability of the administered antibody to reach the sites of tumor. This study was undertaken to use recently developed human monoclonal antibodies to hopefully overcome the problem of antigenicity associated with murine monoclonal antibodies. We have administered either antibody 16.88 or 28A32 to 26 patients and have not noted the development of antibodies to the administered human monoclonal antibody in any patient. Three patients had pre-existing anti-28A32 antibodies, and the titer of antibody increased somewhat in one of these patients. Positive images of tumors were obtained in 20 of the 26 patients; the reason why images were not obtained in the other six patients is unknown, but in one patient in whom an image was not obtained with antibody 28A32, an endogenous antibody with reactivity similar to 28A32 was found. Presumably this antibody acted as a """"""""blocking antibody"""""""" and prevented the administered labeled antibody from reaching the tumor target. The ability of these antibodies to localize to sites of tumor and their lack of immunogenicity in patients with cancer suggest that they might be used to deliver therapeutic doses of radionuclides to sites of tumor.