Surgical excision is an option to diagnose small-sized lung cancer, although this procedure has potential to disseminate tumor cells from the surgical margin. This retrospective study enrolled 252 patients with clinical stage IA non-small cell lung carcinoma who had undergone lobectomy during the period 1998-2004. Except for 25 patients with ground-glass attenuation (GGA) lesions on computed tomography, all underwent preoperative biopsy using flexible fiberoptic bronchoscopy (FFB). A total of 148 patients were diagnosed by FFB, and 86 were diagnosed by surgical excision. In the surgical excision cases, 67 tumors were negative for malignancy at the surgical margins and 19 were positive. Diagnosis by surgical excision was associated significantly more often with smaller tumor size (P<0.0001), a greater number of GGA lesions (P=0.0006) and a lower pathological stage (P=0.001) than those diagnosed by FFB. Furthermore, these patients showed better survival (P=0.03) and fewer local recurrences than patients diagnosed by FFB. In the groups that underwent excision, there was no significant difference in survival between those with positive and negative cytological margins. The survival of patients diagnosed by surgical excision was significantly better than that of those diagnosed by FFB in clinical stage IA disease. Surgical excision is an optimal method to diagnose small lung cancer because the malignant status of the margin does not appear to influence the outcome.