Background: Adjuvant chemotherapy is the standard of care for completely resected stage 2-3 non-small cell lung cancer(NSCLC). A few trials suggest neoadjuvant chemotherapy is a promising mode for resectable NSCLC. Indirect comparison meta-analysis of adjuvant versus neoadjuvant therapy showed no difference in survival. This study was conducted to determine whether neoadjuvant chemotherapy or adjuvant chemotherapy prolongs disease-free survival among patients with resectable NSCLC. Methods: Patients with clinical stage IB-IIIA NSCLC were eligible. Patients were randomly assigned to 3 cycles of neoadjuvant DC (Docetaxel: 75mg/m2, Carboplatin :AUC=5 on day 1 every 3wk),followed by surgery 3-6 wk after chemotherapy, or surgery followed by 3 cycles of adjuvant DC at the same schedule. The primary end point was 3 years Disease Free Survival(DFS); secondary end points were 3 years Overall Survival rate (OS) and Safety. Planned sample size is 410. Results: Between March 2006 and May 2011, 198 patients have been accrued, 97 in the neoadjuvant arm, 101 in the adjuvant arm. The neoadjuvant arm had more patients received chemotherapy( 100% v.s 85.1%, P<0.001 ) and received 3 cycles( 91.8% v.s82.6%, P=0.061) than adjuvant arm. Both arms are well tolerated to DC chemotherapy. The most common grade 3/4 adverse event is neutropenia (41.2% with neoadjuvant arm v.s 31.7% with adjuvant arm). One chemotherapy related death in adjuvant arm. One patient die of perioperative pulmonary embolism in neoadjuvant arm. No difference in peri-operative complication between two arms. The 3 years DFS was 45% in the neoadjuvant arm and 53% in the adjuvant arm, HR=0.88 (0.58-1.33), P=0.54. Median survival has not been reached in both arms. Conclusions: Neoadjuvant or adjuvant chemotherapy with docetaxel plus carboplatin in resectable clinical stage IB-IIIA NSCLC is feasible and safe. Preliminary results show similar 3 years DFS in both arms. The OS data has not matured in both arms. Clinical trial information: NCT00321334.