Docetaxel alone has been confirmed to be beneficial to patients with advanced previously treated non-small cell lung cancer (NSCLC). However, the duration and survival time is short. The study of two-agent combination regimens has important clinical significance. We conducted this randomized controlled phase II trial to comparatively evaluate the efficacy and side effects of capecitabine combined with docetaxel in previously treated patients with NSCLC. Patients with previously treated NSCLC who failed first-line chemotherapy were randomized into two groups; one received capecitabine combined with docetaxel (XT group) and the other received docetaxel alone (T group). Patients in the XT group received chemotherapy as follows: capecitabine 625 mg/m(2), p.o. bid, days 5-18; and docetaxel 30 mg/m(2), days 1 and 8, while patients in the T group received docetaxel 35 mg/m(2) on days I and 8. The primary endpoint was time to progression (TTP), and secondary endpoints were overall survival (OS), response rate (RR) and disease control rate (DCR). Forty-eight patients were recruited (23 in the Xi group and 25 in the T group). TIP, median survival time (MST) and I-year OS rate in the XT group and the T group were 7 months, 12 months, 47.6% and 3 months, 12 months, 39.6%, respectively. The TTP in the XT group was significantly longer compared to that in the T group (chi(2)=-4.763, p=0.029). The RR and DCR in the XT group and T group were 13.0% (3/23), 78.3% (18/23) and 12.0% (3/25), 76% (19/25), respectively. The difference was not significant (p>0.05). The major side effects observed in the two groups were neutropenia, fatigue and nausea, and toxicities were mild to modest. No severe cases of hand-foot syndrome were observed in the XT group. In conclusion, compared with docetaxel alone, capecitabine combined with docetaxel for patients with previously treated NSCLC achieved a significantly longer TIT and this regimen was well tolerated. The relatively high median TTP, 1-year OS rate and DCR encourage further evaluation of this regimen in a randomized phase III trial.