To investigate the chance of discovery of metastatic lung tumors and the five-year survival rates of patients undergoing surgical resection, we followed 99 patients who underwent initial surgical treatment at our hospital between 1979 and 1996. With regard to primary organs or sites, 32 patients had rectal cancer, 27 patients had breast cancer, 19 patients had colon cancer and 21 patients had osteosarcoma. For 22 of 99 patients (22%), discovery was due to subjective symptoms such as cough and sputum (n = 12), chest (or back) pain (n = 7) or hemosputum (n = 5). Ten of 19 patients (53%) with colon cancer experienced subjective symptoms which led to the discovery of metastases. In 76 of 99 patients (78%), metastatic lung lesions were not discovered through subjective symptoms. In 63 of those 76 patients, such lesions were initially found by plain chest roentgenography or CT. In 20 of 21 patients (95%) who had osteosarcoma, metastatic lung tumors were discovered by chest roentgenography or CT. In 14 of 76 patients, all of whom had metastatic lung carcinomas, the lesions were discovered through elevated levels of tumor markers. Therefore the importance of periodic chest roentgenography and tumor marker testing was demonstrated. Disease-free interval (DFI) was over six years in five of 32 patients (16%) with rectal cancer and 13 of 27 (48%) with breast cancer. DFI was less than five years for 15 of 19 patients (79%) with colon cancer, and less than two years for 16 of 21 (75%) with osteosarcoma. Thus, DFI differed according to the sites of the tumors. The five-year survival rates of 97 patients were examined. Patients were divided according to the sites of their primary tumors, and then subdivided according to the type of surgery they received. Patients were thus divided into five categories: I) those who underwent incomplete resection of metastatic lung lesions, II) those who underwent complete resection of both pulmonary lesions and involved mediastinal lymph nodes, III) those who had undergone previous treatment for tumors in organs other than the lung, IV) those who underwent complete resection of multiple lung lesions, and V) those who underwent complete resection of solitary lung lesions. For all primary sites, none of the patients in group I) survived for more than two years. Therefore complete resection seems very important for the treatment of metastatic lung tumors. With regard to the other groups, several facts were noted. For rectal cancer, the five-year survival rate of groups V) and III) was 55.6% in either case. Therefore complete resection of rectal cancer metastatic to the lung may improve the five-year survival rate even for patients who have previously been treated for cancers in organs other than the lung. For colon cancer, the five-year survival rate of group V) was 51.4%. Complete resection of only a solitary lung lesion may improve the five-year survival rate for colon cancer. For breast cancer, the five-year survival rate of group V) was 37.5% and that of group II) was 60.0%. This may indicate that for patients who have both pulmonary lesions and mediastinal lymph node involvement, complete resection of both is important. For osteosarcoma the five-year survival rate of group IV) was 26.0%. Thus, osteosarcoma patients have a chance of survival if they undergo complete resection of lung metastases.