Objective To systematically review the efficacy of preoperative rehabilitation for patients with total knee replacement. Methods Electronic databases including The Cochrane Library, EMbase, PubMed, CBM, CNKI and WanFang Data were searched to collect randomized controlled trials (RCTs) about preoperative rehabilitation for total knee replacement from inception to January 2017. Two reviewers independently screened literature, extracted data, and assessed the risk of bias of included studies. Then RevMan 5.3 software was used to perform meta-analysis. Results A total of 13 RCTs involving 745 patients were included. The results of meta-analysis showed that, the TUG scores (6 weeks: MD=–1.31, 95%CI –2.29 to –0.32, P=0.009; 3 months: MD=–1.59, 95%CI –2.03 to –1.15, P<0.000 01), 3 months knee flexion ROM scores (MD=4.75, 95%CI 0.63 to 8.86, P=0.02) in the preoperative rehabilitation group were higher than those in the control group. The 3 months VAS scores (MD=–0.95, 95%CI –1.70 to –0.20, P=0.01) in the preoperative rehabilitation group was lower than that in the control group. There were no statistical differences between two groups in WOMAC scores (MD=–5.61, 95%CI –12.84 to 1.62, P=0.13), SF-36 scores (MD=13.15, 95%CI –10.51 to 36.81, P=0.28), knee extension ROM scores (6 weeks: MD=0.60, 95%CI –1.02 to 2.21, P=0.47; 3 months: MD=–2.76, 95%CI –8.45 to 2.92, P=0.34), HSS scores (MD=0.54, 95%CI –0.46 to 1.54, P=0.29), and 6 minutes test scores (MD=–8.75, 95%CI –51.53 to 34.03, P=0.69). Conclusion Current evidence shows that preoperative rehabilitation can improve the postoperative knee function after operation and significantly reduce the pain. Due to limited quality and quantity of the included studies, more high-quality studies are needed to verify above conclusion.