Thirteen patients with metastatic malignant melanoma received intravenous therapy with the murine antimelanoma monoclonal antibody 9.2.27. Five patients were entered on a dose escalation protocol with twice weekly escalating doses of 10-500 mg, in an extension of a previously reported trial. These patients demonstrated near saturation of available antibody binding sites in vivo following the 500 mg dose, with minimal toxicity. The remaining patients were entered onto a dose schedule comparison study, with a 500 mg dose administered either in a single 2 h infusion or as five daily 2 h infusions of 100 mg to examine the effects of different dose schedules and of an interrupted schedule on subsequent therapy with the same antibody. Intratumor localization of the monoclonal antibody did not appear to vary with respect to the dose schedule; however, interruption in therapy for 4 weeks was accompanied by somewhat poorer localization of antibody. This effect appeared to be primarily attributable to development of human antimurine antiglobulin in 25-30% of patients with resultant decrease in intratumor localization of antibody and more rapid clearance of the 9.2.27 antibody from the circulation. Earlier reports with other antibodies notwithstanding, initial infusions of 500 mg of 9.2.27 did not induce tolerance to the murine immunoglobulin. This study confirms and extends the findings of our initial trial of the 9.2.27 antibody by demonstrating that, although clinical responses were not observed, the antibody can be safely administered at doses up to 500 mg, with good intratumor localization of antibody. The diminished localization of antibody associated with antiglobulin responses indicates the importance of monitoring antiglobulin levels during therapy, and the necessity of controlling or preventing this phenomenon when monoclonal antibodies are administered in multiple doses as drug, toxin, or radionuclide immunoconjugates.