Between May 1975 and April 1981, 43 adult patients with high-grade soft tissue sarcomas of the extremities were prospectively randomized to receive either amputation at or above the joint proximal to the tumor, including all involved muscle groups, or to receive a limb-sparing resection plus adjuvant radiation therapy. The limb-sparing resection group received wide local excision followed by 5000 rads to the entire anatomic area at risk for local spread and 6000 to 7000 rads to the tumor bed. Both randomization groups received postoperative chemotherapy with doxorubicin (maximum cumulative dose 550 mg/m2), cyclophosphamide, and high-dose methotrexate. Twenty-seven patients randomized to receive limb-sparing resection and radiotherapy, and 16 received amputation (randomization was 2:1). There were four local recurrences in the limb-sparing group and none in the amputation group (p1 = 0.06 generalized Wilcoxon test). However, there were no differences in disease-free survival rates (71% and 78% at five years; p2 = 0.75) or overall survival rates (83% and 88% at five years; p2 = 0.99) between the limb-sparing group and the amputation treatment groups. Multivariate analysis indicated that the only correlate of local recurrence was the final margin of resection. Patients with positive margins of resection had a higher likelihood of local recurrence compared with those with negative margins (p1 less than 0.0001) even when postoperative radiotherapy was used. A simultaneous prospective randomized study of postoperative chemotherapy in 65 patients with high-grade soft-tissue sarcomas of the extremities revealed a marked advantage in patients receiving chemotherapy compared with those without chemotherapy in three-year continuous disease-free (92% vs. 60%; p1 = 0.0008) and overall survival (95% vs. 74%; p1 = 0.04). Thus limb-sparing surgery, radiation therapy, and adjuvant chemotherapy appear capable of successfully treating the great majority of adult patients with soft tissue sarcomas of the extremity.