During a recent 20-year period, 556 patients underwent operation for pulmonary metastasis at the University of Texas M.D. Anderson Hospital and Tumor Institute at Houston. The surgical mortality was 1.5% for 772 resections. A selection of 443 patients was made to evaluate the contribution of operative intervention as a primary treatment, with selective adjunctive therapy when applicable. The success of a surgical approach is dependent primarily on adherence to selection criteria; it is important that only patients in whom all known disease can be completely removed with the planned resection and who have full control of the primary site are treated. The overall survival for the group was 35%. For patients with carcinoma, survival ranged from 24% for those with primary uterine cervix tumors to approximately 54% for urinary tract, male genital tract, and corpus of uterus primary tumors. In the group with sarcoma, patients with skeletal tumors had a 46.4% survival rate (50.7% for those with osteogenic sarcoma), and 33% of the patients with soft tissue tumors had long-term survival. The outcome for patients with melanoma was poor; only 12.1% survived 5 years. If the original criteria apply, multiple and bilateral lesions can be successfully managed. Patients undergoing planned adjuvant treatment had a superior outcome compared with those not so treated. However, a significant survival advantage was shown only for patients with sarcoma. The failure to control all disease in patients in whom pulmonary metastasis is controlled surgically can only be improved on through the use of systemically active adjuvant treatment.