IntroductionKnee osteoarthritis (KOA) is one of the leading causes of disability. The effectiveness of acupuncture for treating KOA remains controversial. This protocol describes the method of a systematic review and meta-analysis evaluating the effectiveness and safety of acupuncture for treating KOA.Methods and analysisFour English databases (PubMed, Embase, Cochrane Library databases and Web of Science) and four Chinese databases (China National Knowledge Infrastructure, Chinese Biomedical Literature Database, VIP Database for Chinese Technical Periodicals, and Wanfang) will be searched from the database inception to 1 September 2021. All randomised controlled trials related to acupuncture for KOA will be included. Extracted data will include publication details, basic information, demographic data, intervention details and patient outcomes. The primary outcome will be pain intensity. Risk of bias will be assessed using the Cochrane Collaboration’s tool for assessing risk of bias. Article selection, data extraction and risk of bias assessment will be performed in duplicate by two independent reviewers. If the meta-analysis is precluded, we will conduct a descriptive synthesis using a best-evidence synthesis approach. The strength of recommendations and quality of evidence will be assessed using the Grading of Recommendations Assessment Development and Evaluation working group methodology.Ethics and disseminationEthics approval is not required because individual patient data are not included. This protocol was registered in the international Prospective Register of Systematic Reviews on 25 February 2021. The systematic review and meta-analysis will be submitted for publication in a peer-reviewed journal. The findings will also be disseminated through conference presentations.Trial registration numberCRD42021232177.