© 2016 Wolters Kluwer Health, Inc. Background: To develop a population pharmacokinetic model of gentamicin in children with complicated severe malnutrition and to study the influence of covariates (weight and age) on pharmacokinetic indices. In addition, we use the model to perform Monte Carlo simulations to explore the efficacy of several dosage regimens. Methods: Twenty-six children with severe complicated malnutrition were studied. Ninety-six samples of gentamicin plasma concentrations, obtained from 0.5 to 8 hours after intravenous dosing, were analyzed. Population pharmacokinetic models were built using the program Monolix 4.2 (Lixoft, Antony, France). Monte Carlo simulations were performed to evaluate optimal dosage regimens, using the final pharmacokinetic model, based on the probability of pharmacokinetic-pharmacodynamic target attainment. Results: The concentration-time data were fitted best to 1-compartment model. The estimated population clearance was 1.1 L/h, and the volume of distribution was 2.23 L, with an interindividual variability of 47.2% and 35.6%, respectively. The final models for the clearance and volume of distribution were as follows: CL (L/h) = CL = 1.15 (age/median age)0.321 and V (L) = 2.33 (weight/median weight)0.743. In Monte Carlo simulations, gentamicin given in dosages of 7.5 to 15 mg/kg optical density was effective in achieving the pharmacodynamic target Cmax:minimal inhibitory concentration >10 for minimal inhibitory concentrations ≤2.5 mg/L, with a probability lower than 1% for Cmin >1 mg/L. Conclusions: Based on the available evidence, an intravenous dose of 7.5 to 15 mg/kg once daily in children with complicated severe malnutrition and normal renal function ensures high probability of efficacy and low risk of nephrotoxicity, which gives further support to the recommendations issued by the World Health Organization treatment for this patient population.