Objective: To compare the effect of three kinds of traditional Chinese medicine injection (zhengqing fengtongning injection, shenmai injection, cervus and cucumis polypeptide injection) on knee osteoarthritis with the method of Network meta-analysis. Methods: Two researchers independently used computers to search for randomized controlled clinical trials of zhengqing fengtongning injection, shenmai injection, or cervus and cucumis polypeptide injection for the treatment of knee osteoarthritis published in CNKI,VIP,WANFANG DATA,CBM and PubMed databases (since the date of database establishment to February 2020). After the methodological quality of included researches were evaluated, and the valid data were extracted, statistical analysis was performed With RevMan5.3 and R 3.53. Results: There were including 9 RCTs, which involving 720 patients.The results of network Meta-analysis showed that: All three traditional Chinese medicine injections effects are approximately equal. The probability ranking chart shows that Zhengqingfengtongning injection may have the best effect. Conclusions: Three commonly used traditional Chinese medicine injections can effectively treat knee osteoarthritis, of which Zhengqingfengtongning injection is probably the most effective. Due to the small number of literatures included in this study and the single outcome index, we still need to treat the results of this study with caution. It is suggested that more high-quality, long-term follow-up studies should be carried out for this kind of treatment in the future, so as to provide solid evidence support for the popularization of these treatment measures.