Video-assisted thoracic surgery (VATS) lobectomy for early lung cancer has become technically feasible. We sought to determine if VATS preserved chest wall muscle postoperatively better than thoracotomy. Consecutive patients who underwent lobectomy between 2004 and 2006 for clinical Stage IA non-small cell lung cancer through VATS (VATS group) or posterolateral thoracotomy (PLT group) at our institution were eligible for the study. The cross-sectional areas of bilateral latissimus dorsi muscle (LDM) at the lower end of the scapula were obtained by computed tomography preoperatively and one year after surgery. These were quantified with image analysis by two researchers in a blinded manner. Fourteen patients in the VATS group (mean age, 68 years; 8 men, 6 women) and 24 patients in the PLT group (mean age, 62 years; 14 men, 10 women) were assessed. Postoperative/preoperative ratios of the LDM cross-section areas on the surgical side were 89+/-20% (Mean+/-S.D.) in the VATS group and 57+/-16% in the PLT group (P<0.001). Those on the non-surgical side were 89+/-23% in the VATS group and 97+/-16% in the PLT group (P=0.23). We conclude that VATS may prevent atrophy of LDM on the surgical side better than conventional thoracotomy.