Objective: Malnutrition is a major cause of childhood morbidity and mortality. To identify and target those at highest risk, there is a critical need to characterize biomarkers that predict complications prior to and during treatment. Methods: We used targeted and nontargeted metabolomic analysis to characterize changes in a broad array of hormones, cytokines, growth factors, and metabolites during treatment of severe childhood malnutrition. Children aged 6 months to 5 years were studied at presentation to Mulago Hospital and during inpatient therapy with milk-based formulas and outpatient supplementation with ready-to-use food. We assessed the relationship between baseline hormone and metabolite levels and subsequent mortality. Results: Seventy-seven patients were enrolled in the study; a subset was followed up from inpatient treatment to the outpatient clinic. Inpatient and outpatient therapies increased weight/height z scores and induced striking changes in the levels of fatty acids, amino acids, acylcarnitines, inflammatory cytokines, and various hormones including leptin, insulin, GH, ghrelin, cortisol, IGF-I, glucagon-like peptide-1, and peptide YY. A total of 12.2% of the patients died during hospitalization; the major biochemical factor predicting mortality was a low level of leptin (P = .0002), a marker of adipose tissue reserve and a critical modulator of immune function. Conclusions: We have used metabolomic analysis to provide a comprehensive hormonal and metabolic profile of severely malnourished children at presentation and during nutritional rehabilitation. Our findings suggest that fatty acid metabolism plays a central role in the adaptation to acute malnutrition and that low levels of the adipose tissue hormone leptin associate with, and may predict, mortality prior to and during treatment.