OBJECTIVES: To compare the efficacy and safety of intravenous rehydration (IVR) versus oral rehydration (OR) strategies in children with severe acute malnutrition (SAM) and severe dehydration. DESIGN: Systematic review and meta-analysis. SETTING: Lower-to-middle-income countries. PATIENTS: Children with SAM are hospitalised with severe dehydration secondary to gastroenteritis. INTERVENTIONS: Randomised trials of IVR and OR (standard of care) MAIN OUTCOME MEASURES: Primary: in-hospital mortality. Secondary: fluid overload events, development of shock requiring intravenous boluses, severe electrolyte abnormalities at 24 hours and day 28 mortality. RESULTS: We identified three RCTs comprising 484 participants with severe malnutrition (72 had kwashiorkor, two had some risk of bias, and one had low risk of bias). The pooled risk ratio (RR) for in-hospital mortality for IVR versus OR is 0.71 (95% CI 0.46 to 1.10; I2=0.0%) with moderate certainty of evidence. No fluid overload events were reported, pooled RR 0.99 (95% CI 0.10 to 9.35). The pooled RR of severe hyponatraemia (sodium <125 or <130 mmol/L) at 24 hours was 0.66 (95% CI 0.44 to 0.99). Only one trial reported RR for shock development, hypernatraemia (sodium >145 mmol/L) or 28-day mortality with IVR versus OR (RRs 0.56, 95% CI 0.21 to 1.48; RR 2.05, 95% CI 0.50 to 8.58 and RR 0.85, 95% CI 0.44 to 1.65, respectively). Subgroup analyses for in-hospital mortality were carried out for region and risk of bias rating, giving p=0.85 and p=0.54 for heterogeneity, respectively. CONCLUSIONS: The estimated effect of using IVR versus OR in children with SAM with severe dehydration ranges from a 54% relative reduction to a 10% relative increase in the risk of death, with IVR resulting in fewer adverse events. PROSPERO REGISTRATION NUMBER: CRD42025637956.