BackgroundRecombinant human endostatin (rh‐endostatin) plus standard chemotherapy in advanced non‐small cell lung cancer (NSCLC) patients has shown improved efficacy; however, it is unclear whether it is effective and safe when added to pemetrexed/cisplatin and used as maintenance therapy.MethodsWe retrospectively evaluated the data of untreated NSCLC patients administered rh‐endostatin plus pemetrexed/cisplatin or pemetrexed/cisplatin. The primary endpoint was progression‐free survival (PFS).ResultsFifty‐six and 39 patients received rh‐endostatin plus pemetrexed/cisplatin and pemetrexed/cisplatin, and 34 and 29 underwent maintenance treatment, respectively. The median PFS was 10 months (95% confidence interval [CI] 5.85–14.15) in the rh‐endostatin and 8.2 months (4.04–12.36) in the chemotherapy group, but the difference was not statistically significant (P = 0.13). In patients administered maintenance treatment, rh‐endostatin plus pemetrexed was associated with prolonged PFS compared to single‐agent pemetrexed when PFS was calculated from first dosing (13.7 [9.41–17.99] vs. 8.2 [4.16–12.24]; P = 0.032); however, PFS did not differ between the groups (hazard ratio 0.618; 95% CI 0.368–1.038; P = 0.069) after adjusting for clinical factors. No difference was observed in the objective response rate between the groups (48.2% vs. 38.5%; P = 0.346), with the exception of men (62.1% vs. 33.3%; P = 0.032) or in the incidence of drug‐related or grade 3–4 adverse events.ConclusionIn previously untreated, advanced‐stage NSCLC patients, first‐line treatment with pemetrexed/cisplatin plus rh‐endostatin did not prolong PFS or overall survival when compared to pemetrexed/cisplatin, but a trend of improved PFS was observed in patients administered maintenance rh‐endostatin plus pemetrexed.