OBJECTIVE: Cupping therapy is widely used in East Asia, the Middle East, or Central and North Europe to manage the symptom of knee osteoarthritis (KOA). The purpose of this systematic review was to evaluate the available evidence from randomized controlled trials (RCTs) of cupping therapy for treating patients with KOA. METHODS: The following databases were searched from their inception until January 2017: PubMed, Embase, the Cochrane Central Register of Controlled Trials and four Chinese databases [WanFang Med Database, Chinese BioMedical Database, Chinese WeiPu Database, and China National Knowledge Infrastructure (CNKI)]. Only the RCTs related to the effects of cupping therapy on KOA were included in this systematic review. A quantitative synthesis of RCTs will be conducted using RevMan 5.3 software. Study selection, data extraction, and validation was performed independently by two reviewers. Cochrane criteria for risk-of-bias were used to assess the methodological quality of the trials. RESULTS: Seven RCTs met the inclusion criteria, and most were of low methodological quality. Study participants in the dry cupping therapy plus the Western medicine therapy group showed significantly greater improvements in the pain [MD = -1.01, 95%CI (-1.61, -0.41), p < 0.01], stiffness [MD = -0.81, 95%CI (-1.14, -0.48), p < 0.01] and physical function [MD = -5.53, 95%CI (-8.58, -2.47), p < 0.01] domains of Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) compared to participants in the Western medicine therapy group, with low heterogeneity (Chi = 0.00 p = 1.00, I = 0% in pain; Chi = 0.45 p = 0.50, I = 0% in stiffness; Chi = 1.09 p = 0.30, I = 9% in physical function). However, it failed to do so on a Visual Analog Scale (VAS) [MD = -0.32, 95%CI (-0.70, 0.05), p = 0.09]. In addition, when compared with Western medicine therapy alone, meta-analysis of four RCTs suggested favorable statistically significant effects of wet cupping therapy plus western medicine on response rate [MD = 1.06, 95%CI (1.01, 1.12), p = 0.03; heterogeneity: Chi = 1.13, p = 0.77, I = 0%] and Lequesne Algofunctional Index (LAI) [MD = -2.74, 95%CI (-3.41, -2.07), p < 0.01; heterogeneity: Chi = 2.03, p = 0.57, I = 0% ]. CONCLUSION: Only weak evidence can support the hypothesis that cupping therapy can effectively improve the treatment efficacy and physical function in patients with KOA.