During the past 10 years, (November, 1973 through November, 1983) 208 patients with fully malignant primary osteogenic sarcoma of an extremity were treated with preoperative chemotherapy on 4 successive treatment protocols. Continuous improvements in the disease-free survival of patients were attributed to refinements in the chemotherapy regimens. These refinements were made after direct observation of the response of the primary tumor to chemotherapy. At a minimum follow-up time of over 38 months for 87 patients treated on the T-10 chemotherapy protocol, 67 (77%) have remained alive and continuously free of disease, and 71/87 (81.6%) are currently free of disease at a median follow-up time of 5 years. Overall complete response rate of the primary tumor to preoperative chemotherapy was 48%. Fifty-one patients were treated on a pilot protocol (T-12) from November, 1981, to November, 1983. The main difference was that after preoperative high dose methotrexate and combination bleomycin, cytoxan and dactinomycin therapy, patients having a good histologic response of the primary tumor (21/51 or 41%) had their chemotherapy stopped at 15 weeks and did not receive platinum or Adriamycin chemotherapy. 38/51 (75%) on T-12 remained continuously free of disease and 39/51 (76%) are currently alive and free of disease. There were 2 local recurrences on the T-10 protocol and 6 on the T-12 protocol. Excluding local recurrences 71/85 (84%), patients treated on T-10 are currently alive and free of disease and 38/45 (84%) patients treated on T-12 are alive and free of disease (at a median follow-up of 24 months for T-12). This preliminary study indicates that preoperative chemotherapy may be used to select a subset of patients (who respond well to preoperative high dose methotrexate) who can have therapy terminated early, sparing those patients the undesirable side effects and cost of additional therapy with platinum and Adriamycin.