Background Several clinical trials showed that erlotinib was effective after the failure of gefitinib in advanced non-small cell lung cancer (NSCLC). The aim of this study was to evaluate the feasibility of erlotinib treatment after the failure of gefitinib based on the data from our hospital.Methods The clinical data of 20 patients with advanced NSCLC who were admitted to Shanghai Chest Hospital from August 2007 to December 2008 were retrospectively analyzed. All of the patients were given erlotinib treatment after the failure of gefitinib. Survival analysis was made by Kaplan-Meier method. The Cox regression model was performed to analyze the relationship between the influential factors and the erlotinib progression-free survival (PFS).Results Five patients had a partial response (PR), nine patients had stable disease (SD) and six patients had progressive disease (PD) with gefitinib treatment. The median PFS was 277 days (95% C/0-566). No patient had a PR,seven had SD and fourteen PD with the erlotinib therapy. The median PFS was 31 days (95% Cl 9.1-52.9). The response rate (RR) was 0, and the disease control rate (DCR) was 35% (7/20). Cox regression analysis demonstrated that sex (P=0.96), age (P=0.89), smoking history (P=0.78), performance status (PS) (P=0.98), gefitinib efficacy (P=0.90)and whether chemotherapy was applied between using the two drugs (P=0.45) had no significant correlation with erlotinib PFS. Fifteen patients had epidermal growth factor receptor (EGFR) mutation status determined. There were five cases got SD with the erlotinib treatment in ten mutation negative (wild-type) patients. No SD was recorded in the five mutation positive patients.Conclusions The efficacy of erlotinib treatment after gefitinib failure was limited. However, the patients who are EGFR mutation negative can probably benefit from erlotinib treatment after gefitinib failure.