The aim of this meta-analysis was to evaluate the preventive efficacy and safety of 5-HT3 receptor antagonists (5-HT3 RAs) on perioperative shivering. Relevant databases were searched to identify eligible randomized, controlled trials through January 2016. Primary outcome was the incidence of perioperative shivering, and secondary outcomes were the incidence of safety-related outcomes including postoperative nausea and vomiting (PONV), bradycardia, and hypotension. We calculated risk ratios (RRs) with 95% confidence intervals (CIs) for dichotomous data. Trial sequential analysis was performed to assess the risk of random errors and calculate the required information size. Sixteen studies with a total of 1126 patients were included in the meta-analysis. Compared with the control group, 5-HT3 RAs administered intravenously could statistically significantly reduce the incidence of perioperative shivering (RR, 0.44; 95%CI, 0.35 to 0.56; P < .00001; heterogeneity: I2 = 30%) as well as PONV (RR, 0.52; 95%CI, 0.28 to 0.97; P = .04; heterogeneity: I2 = 0%). However, they did not show superiority in lowering the rate of bradycardia (RR, 0.75; 95%CI, 0.38 to 1.49; P = 0.42; heterogeneity: I2 = 0%) or hypotension (RR, 0.79; 95%CI, 0.44 to 1.43; P = .44; heterogeneity: I2 = 24%). Trial sequential analysis of primary outcome showed that the required information size was 2634 patients and that the trial sequential monitoring boundary was crossed. Thus, more high-quality randomized, controlled trials with larger sample sizes are still required to draw a definite conclusion about the preventive efficacy of 5-HT3 RAs on perioperative shivering prevention in the future.