Objective:To analyze the efficacy and safety of curcumin in the treatment of knee osteoarthritis.Methods:The randomized controlled trials of curcumin in the treatment of knee osteoarthritis published from January 2011 to August 2021 were retrieved. The bias risk of the included literatures was evaluated by Revman 5.3 software, and the efficacy related indexes and the incidence of adverse events were analyzed by Stata 16.0 software. The weighted mean difference ( WMD) was calculated for the difference of efficacy indexes, the odds ratio ( OR) was calculated for the difference of safety indexes, the difference was compared by t test. Results:① A total of 9 relevant literatures were included, all of which were in English. ② A total of 724 patients were included in the study, of which 383 were treated with curcumin capsules and 341 were treated with placebo. ③ The visual analogue scale/score (VAS) of patients treated with oral curcumin at 3-4, 6 and 8 weeks were significantly lower than those of patients treated with oral placebo, the differences were statistically significant [weighted mean difference ( WMD)=-1.09, 95% CI (-1.44, -0.73), P<0.001; WMD=-1.52, 95% CI (-2.35, -0.69), P<0.001; WMD=-1.20, 95% CI(-1.71, -0.69), P<0.001]. ④ The western Ontario and McMaster universities osteoarthritis index (WOMAC) scores of patients treated with oral curcumin for 3-4 and 6-8 weeks were significantly lower than those of patients treated with oral placebo, and the differences were statistically significant [ WMD=-7.96, 95% CI(-14.89, -1.04), P=0.020; WMD=-15.34, 95% CI(-20.51, -10.18), P<0.001]. Specifically, the WOMAC pain and stiffness scores of patients treated with oral curcumin for 6-8 weeks were significantly lower than those of patients treated with oral placebo, and the differences were statistically significant [ WMD=-2.16, 95% CI(-3.69, -0.63), P=0.010; WMD=-1.00, 95% CI (-1.54, -0.46), P<0.001]. The WOMAC joint function scores of patients treated with oral curcumin for 3-4 and 6-8 weeks were significantly lower than those of patients treated with oral placebo, the difference was statistically significant [ WMD=-3.21, 95% CI(-4.51, -1.92), P<0.001; WMD=-7.07, 95% CI(-11.19, -2.94), P<0.001]. ⑤ There was no significant difference in the incidence of adverse events between oral curcumin and placebo [ OR=1.19, 95% P(0.74, 1.90), P=0.478]. Conclusion:Compared with placebo, oral curcumin can significantly alleviate the pain, stiffness and joint function of patients with knee osteoarthritis, and its safety is similar to placebo.