OBJECTIVE Adjuvant chemotherapy improves disease-free survival (DFS) for patients with osteogenic sarcoma (OS). We reviewed our experience with OS to determine prognostic factors, the role of preoperative chemotherapy and subsequent histologic response, and the role of salvage chemotherapy after poor initial response. METHODS From 1975 to 1984, we saw 279 patients with previously untreated OS without metastasis. All patients received intensive chemotherapy and underwent surgical resection of primary tumor. Chemotherapy included high-dose methotrexate; Adriamycin (doxorubicin; Adria Laboratories, Columbus, OH); and bleomycin, cyclophosphamide, and dactinomycin (BCD). Selected patients also received cisplatin. RESULTS DFS was not affected by use of preoperative chemotherapy versus immediate surgery, by use of limb-sparing surgery versus amputation, age, sex, or dose intensity of chemotherapy. DFS did correlate with serum lactate dehydrogenase (LDH), alkaline phosphatase, primary tumor site, race, and histologic response to preoperative chemotherapy. There was no difference in DFS for patients with a poor histologic response who did or did not receive cisplatin, although patients who did receive cisplatin had a longer time to relapse. The 5-year DFS was 76% for patients aged less than or equal to 21 years who had extremity primary tumor and were treated with the T10 protocol. CONCLUSIONS Intensive chemotherapy can achieve DFS for a high proportion of patients with OS. Although it is a powerful predictor of DFS, histologic response to preoperative chemotherapy cannot be assessed at diagnosis. We have not shown an ability to salvage patients with an unfavorable response. We need to increase the proportion of patients with a favorable response, identify the patients who will have an unfavorable response, and develop novel treatments to salvage poor responders.