Objective: To synthesize evidence on the effects of blood flow restriction (BFR) comparing with high (HLT) and low load (LLT), and on the influence of different forms of application in individuals with knee osteoarthritis. Data sources: The CENTRAL, PEDro, PubMed and BVS, which include Lilacs, Medline and SciELO, until April 2020. Review methods: A systematic review and meta-analysis of randomized trials used the PRISMA guidelines, whose main keywords were: Therapeutic Occlusion, Resistance Training, and Knee Osteoarthritie, blood flow restriction and Kaatsu training. Method quality was evaluated with the PEDro scale. When studies demonstrated homogeneity on outcome measures, the mean differences or standardized mean differences with 95% confidence interval were calculated and pooled in a meta-analysis for pooled synthesis. Results: Five articles were eligible in this review with moderate to low risk bias. Our results, showed no difference between BFR and HLT in knee strength (SMD = 0.00, 95% CI, –0.54 to 0.54, P = 1.00), function (SMD = −0.20, 95% CI, –0.45 to 0.06, P = 0.13), pain and volume. But, when compared BFR and LLT, the descriptive analysis demonstrated significant results in favor BFR to muscle strength (71.4% of measurement) and volume (MD = 1.66, 95% CI, 0.93 to 2.38, P < 0.00001), but not in pain or function. Conclusion: BFR can be used as a strategy in the rehabilitation of osteoarthritis due to gains in strength and volume with low mechanical stress. However, its application must be safe and individualized, since they can attenuate the stimuli offered by BFR.