BackgroundScaling up improved management of severe acute malnutrition (SAM) has been identified as the nutrition intervention with the greatest potential to reduce child mortality but it requires improved operational capacity.ObjectiveTo investigate whether an eLearning course, which can be used at scale in resource-poor countries, leads to improved diagnosis, clinical management and survival of children with SAM.DesignA 2-year preintervention and postintervention study between January 2015 and February 2017.SettingEleven healthcare facilities: nine in Ghana, one in Guatemala, and one in El Salvador.InterventionScenario-based eLearning course ‘Caring for infants and young children with severe malnutrition’.Main outcome measuresIdentification of children with SAM, quality of care, case-fatality rate.MethodsMedical record reviews of children aged 0–60 months attending eleven hospitals between August 2014 and July 2016, observations in paediatric wards, and interviews with senior hospital personnel.ResultsPostintervention there was a significant improvement in the identification of SAM: more children had the requisite anthropometric data (34.9% (1300/3723) vs 15.9% (629/3953)) and more were correctly diagnosed (58.5% (460/786) vs 47.1% (209/444)). Improvements were observed in almost all aspects of the WHO ‘Ten Steps’ of case-management, and case-fatality fell from 5.8% (26/449) to 1.9% (14/745) (Post-pre difference=−3.9%, 95% CI −6.6 to −1.7, p<0.001).ConclusionsHigh quality, interactive eLearning can be an effective intervention in scaling up capacity building of health professionals to manage SAM effectively, leading to a reduction in mortality.