After 10 years of clinical trials in patients with advanced cancer, monoclonal antibodies (mAbs) against cell surface antigens have not lived up to their initial promise. One such cell surface antigen is the ganglioside GD2. GD2 is richly expressed at the cell surfaces of human neuroblastomas, sarcomas, and melanomas. We have described a murine lymphoma (EL4) that is syngeneic in C57BL/6 mice and expresses GD2, a mAb against GD2 (mAb 3F8), and we have prepared a conjugate vaccine (GD2-keyhole limpet hemocyanin plus immunological adjuvant QS-21) that consistently induces antibodies against GD2. We demonstrate here, for the first time in a syngeneic murine model, that passively administered and vaccine-induced antiganglioside antibodies prevent outgrowth of micrometastases, and we use this model to establish some of the parameters of this protection. The level of protection was proportional to antibody titer. Treatment regimens resulting in the highest titer antibodies induced the most protection, and protection was demonstrated even when immunization was initiated after tumor challenge. Treatment with 3F8 1, 2, or 4 days after i.v. tumor challenge was highly protective, but waiting until 7 or 10 days after challenge resulted in minimal protection. The results were similar whether number of liver metastases or survival was used as the end point. These results suggest that unmodified mAbs or antibody-inducing vaccines against GD2 (and possibly other cancer cell surface antigens) should be used exclusively in the adjuvant setting, where circulating tumor cells and micrometastases are the primary targets.