A prospective randomized trial (the second cooperative study) was conducted from July 1985 to December 1987 to investigate the benefits of postoperative adjuvant chemotherapy in patients for whom non-small cell lung cancer had been resected completely. Patients were randomly assigned either to a chemotherapy group (group A) treated postoperatively with CDDP (66 mg/m2 x 1), ADM (26 mg/m2 x 1) and UFT (8 mg/kg/day) during 6 months, or to a control group (group B) which had undergone surgery only. Three hundred and thirty-three resected cases were registered. Among them, 24 cases (7.2%) were excluded, because of incomplete resection (15), pathologically benign tumour (3), small cell lung cancer (2) and other factors (4). Three hundred and nine cases were eligible: 155 cases in group A (p-Stage I 93, II 19, III 43) and 154 in group B (I 109, II 10, III 35). The 5-year survival rate in group A was 61.8%, and that in group B 58.1%. The 5-year disease-free survival rate for each group was 61.8% and 57.4%, respectively. There were no significant differences in the 5-year survival between the two groups. However, since a significant difference was observed between the two groups regarding pathological lymph node metastasis (pN), the prognostic factors were adjusted using Cox's proportional hazard model. Thereafter the adjusted survival rate and disease-free survival rate for group A became significantly higher than for group B (P = 0.044 and P = 0.036, respectively). Thus, from these results, it is concluded that the role of surgery for non-small cell lung cancer still remains of primary importance, and postoperative adjuvant chemotherapy is effective to improve the results of surgery and prolong life of patients with non-small cell lung cancer.