OBJECTIVE: The tourniquet is currently widely used in total knee arthroplasty. However, whether the application of tourniquet can increase the thickness of bone cement around the prosthesis, thereby increasing the stability and long-term survival of the prosthesis, is still unclear. This article intends to determine whether the use of tourniquet will increase the penetration thickness of bone cement through meta-analysis. METHODS: Cochrane Library, Medline, Embase, PubMed, CNKI Chinese database, Wanfang database and other databases were retrieved to search the randomized controlled trials of whether the tourniquet was used in total knee arthroplasty from inception to April 21, 2020. The test that met the standard was evaluated using Cochrane system evaluation method, and meta-analysis was performed using RevMan 5.2 software. The main outcome indicators were bone cement penetration thickness and visual analogue scale score; and the secondary outcome indicators were blood loss, blood transfusion rate, postoperative hemoglobin reduction, and complications. RESULTS: (1) Finally, 11 randomized controlled trials were included. A total of 863 patients were included, containing 432 patients with tourniquet and 431 patients without tourniquet in total knee arthroplasty. (2) Meta-analysis results showed that intraoperative use of tourniquet could effectively increase the bone cement penetration thickness around the tibial part of the prosthesis (MD=0.22, 95%CI:0.12-0.31, P < 0.000 01), but increased the incidence of complications (OR=4.02, 95%CI:2.11-7.67, P < 0.001), increased postoperative pain (MD=1.14, 95%CI:0.30-1.98, P=0.008). Reducing blood loss, blood transfusion rate and postoperative hemoglobin reduction had no significant significance (P > 0.05). CONCLUSION: The application of tourniquet in total knee arthroplasty can increase the thickness of bone cement around the tibial part of the prosthesis, and may increase the long-term survival rate of the prosthesis. However, limited to the current literature quality and sample size, the above conclusions need more high-quality research to verify. [ABSTRACT FROM AUTHOR]