Objective To study the inhibitory effect of recombinant human endostatin injection-endostar against lymphatic metastasis of non-small cell lung cancer (NSCLC). Methods The mouse tumor model was established by subcutaneously injecting 0.2 mL A549 human lung cancer cell suspension (1×10⁷/mL) in the right armpit of nude mice. Fifty-six tumor-bearing nude mice were randomly divided into eight groups(n=7), including control group (physiological saline 0.2 mL,1/d), cisplatin group (cisplatin 20 μg, every other day), three endostar groups (endostar 2, 4, 6 mg/kg;1/d), and three endostar plus cisplatin groups (adding cisplatin 20 μg to three endostar groups, every other day). The drugs were injected via the tail vein for 14 d, and the animals were observed for 7 d following the last administration. Tumor diameters were recorded in each group and the changes of tumor size were observed. The tumor tissues and suspicious lymph nodes were subjected to H-E staining and immunohistochemical staining; the staining results were used to calculate the lymphatic metastasis rate, number of metastatic lymph nodes and micro-lymphatic vessel density (MLVD) of each group. Results The pre- and post-treatment differences of tumor diameters in all the experimental groups were significantly smaller than that in the control group, and those of the three endostar plus cisplatin groups were significantly smaller than those of the three endostar groups or cisplatin group (P<0.05). The lymphatic metastasis rate, number of metastatic lymph nodes and MLVD in three endostar groups and three endostar plus cisplatin groups were significantly less than that in control group and cisplatin group (P<0.05). Lymphatic metastasis positive rate, number of metastatic lymph nodes and MLVD in endostar plus cisplatin group were significantly lower or smaller than those in the control group and cisplatin group(P<0.05); for the three endostar groups and the three endostar plus cisplatin groups, the above three parameters in 6 mg/kg endostar group and 6 mg/kg endostar plus cisplatin group were significantly lower or smaller than those in 2 mg/kg and 4 mg/kg endostar groups or 2 mg/kg and 4 mg/kg endorstar plus cisplatin groups (P<0.05). Conclusion Both endostar and cisplatin can inhibit tumor growth, and combination of both has a stronger inhibition than they are used alone. Endostar can inhibit tumor lymphangiogenesis and lymphatic metastases in a concentration associated manner, which is not found for cisplatin.