We analyzed the cases of twenty-three patients who had Stage-IIB osteosarcoma and skip metastasis to determine the rates of disease-free and long-term survival. The regimens of preoperative and postoperative chemotherapy varied. The patterns of relapse and long-term survival were studied in relation to the skip lesions, and these patterns were compared with those of 224 patients who had Stage-II osteosarcoma but no skip lesion. Of the twenty-three patients who had a skip lesion, twenty-two had either a local recurrence or a distant metastasis; twenty-two patients died, and one remained disease-free at thirty-eight months. Kaplan-Meier analysis showed significant differences in the rates of local recurrence and distant metastasis (Mantel-Cox test statistic, p less than 0.0001) and in the over-all survival (Mantel-Cox test statistic, p less than 0.0001) between the patients who had and those who did not have skip metastasis. The cases of fourteen patients who had skip metastasis from a lesion in the distal end of the femur were compared with those of eighty-seven patients who had a similarly situated primary lesion but no skip metastasis. The difference was significant, although less so than when lesions in all anatomical sites were analyzed. The follow-up data indicated that the use of adjuvant chemotherapy did not improve the poor prognosis of patients who had skip metastasis. Therefore, we regard a skip lesion, along with regional lymphatic metastasis or metastasis to a distant organ, as another criterion for considering an osteosarcoma to be Stage III.