Data from a series of 569 patients with 'curative' resection of non-oat cell tumors were analyzed by the life-table method to evaluate the validity of the postresection pathological staging classification suggested by the American Joint Committee. The cell types were as follows: squamous, 305; adenocarcinoma, 172; large cell, 73; and mixed, 19. Each patient was assigned a pathological TN classification on examination of the resected specimen (all patients were judged clinically to have no distant metastases - MO). There were 173 lesions classified as T1 No. 37 as T1 N1, 212 as T2 NO, 115 as T2, N1, and 32 as either T3 with any N or N2 with any T. In analyzing the data, we identified a subset of lesions (25), initially staged as T2 NO, which were small central lesions, 3 cm or less, located distal to a lobar takeoff. Regardless of the presence of atelectasis or pneumonitis to the hilar area, patients with these lesions had a survival rate similar to that of patients with lesions larger than 3 cm, which could be classified as T2 NO regardless of their location. When lymph nodes were affected (N1), patients with small central lesions (20) had a better survival rate than the patients with either T1 N1 or other T2 N1 lesions. It is therefore suggested that all small central lesions, 3 cm or less, distal to a lobar takeoff be considered T1 lesions. Patients with T1 N1 lesions had a 3 year survival rate of only 36.7%, which is similar to the 39.8% 3-year survival rate of those with T2 N1 lesions. The other patients in Stage 1 had a much better survival rate: Patients with T1 NO lesions had 3 and 5 year survival rates of 68.5% and 54.4%, and those with T2 NO lesions, 53.6% and 40.0%, respectively. Therefore, it would appear more appropriate to classify these patients as having Stage II rather than Stage I disease.