AbstractAimTo review the evidence of the effects of neonatal magnesium sulphate for neuroprotection in perinatal asphyxia and hypoxic‐ischaemic encephalopathy (HIE).MethodThis was a systematic review of randomized controlled trials (RCTs) (with meta‐analysis) and non‐RCTs assessing magnesium sulphate for treating perinatal asphyxia and HIE at 35 weeks or more gestation (primary outcomes: neonatal death and death or long‐term major neurodevelopmental disability).ResultsTwenty‐five RCTs (2099 infants) and four non‐RCTs (871 infants) were included, 23 in low‐ and middle‐income countries (LMICs). In RCTs, reductions in neonatal death with magnesium sulphate versus placebo or no treatment (risk ratio [RR] = 0.68; 95% confidence interval [CI] = 0.53–0.86; 13 RCTs), and magnesium sulphate with melatonin versus melatonin alone (RR = 0.74; 95% CI = 0.58–0.95; one RCT) were observed. No difference in neonatal death was seen for magnesium sulphate with therapeutic hypothermia versus therapeutic hypothermia alone (RR = 0.66, 95% CI = 0.34–1.26; three RCTs), or magnesium sulphate versus phenobarbital (RR = 3.00; 95% CI = 0.86–10.46; one RCT). No reduction in death or long‐term neurodevelopmental disability (RR = 0.52; 95% CI = 0.14–1.89; one RCT) but reductions in several short‐term adverse outcomes were observed with magnesium sulphate. Evidence was low‐ to very‐low certainty because of risk of bias and imprecision.InterpretationGiven the uncertainty of the current evidence, further robust neonatal magnesium sulphate research is justified. This may include high‐quality studies to determine stand‐alone effects in LMICs and effects with and after therapeutic hypothermia in high‐income countries.