We conducted both conventional pairwise and Bayesian network meta-analyses to compare the clinical properties of supraglottic airway devices in children. We searched six databases for randomised clinical trials. Our primary end-points were oropharyngeal leak pressure, risk of insertion failure at first attempt, and blood staining risk. The risk of device failure, defined as the abandonment of the supraglottic airway device and replacement with a tracheal tube or another device, was also analysed. Sixty-five randomised clinical trials with 5823 participants were identified, involving 16 types of supraglottic airway device. Network meta-analysis showed that the i-gel™, Cobra perilaryngeal airway™ and Proseal laryngeal mask airway (LMA -Proseal) showed statistically significant differences in oropharyngeal leak pressure compared with the LMA -Classic, with mean differences (95% credible interval, CrI) of 3.6 (1.9-5.8), 4.6 (1.7-7.6) and 3.4 (2.0-4.8) cmH O, respectively. The i-gel was the only device that significantly reduced the risk of blood staining of the device compared with the LMA-Classic, with an odds ratio (95%CrI) of 0.46 (0.22-0.90). The risk (95%CI) of device failure with the LMA-Classic, LMA -Unique and LMA-Proseal was 0.36% (0.14-0.92%), 0.49% (0.13-1.8%) and 0.50% (0.23-1.1%), respectively, whereas the risk (95%CI) of the i-gel and PRO-Breathe was higher, at 3.4% (2.5-4.7%) and 6.0% (2.8-12.5%), respectively. The risk, expressed as odds ratio (95%CrI), of insertion failure at first attempt, was higher in patients weighing < 10 kg at 5.1 (1.6-20.1). We conclude that the LMA-Proseal may be the best supraglottic airway device for children as it has a high oropharyngeal leak pressure and a low risk of insertion. Although the i-gel has a high oropharyngeal leak pressure and low risk of blood staining of the device, the risk of device failure should be evaluated before its routine use can be recommended.