Colistin is increasingly used as a salvage therapy for nosocomial infections caused by multidrug-resistant Gram-negative bacteria such as Pseudomonas aeruginosa and Acinetobacter baumannii. However, the available pharmacokinetic (PK) data for colistin are limited to guide dosing. The aim of this study was to develop a population PK model of colistin and to identify the optimal dosage regimens for burn patients. Fifty patients with burns ranging from 4% to 85% of total body surface area who had been treated with colistimethate sodium (CMS) were studied. CMS, which is hydrolyzed in vivo to an active metabolite, was intravenously administered every 12 h. Blood samples were collected at 0, 1, 2, 4, 6, and 8 h after more than five infusions to measure the colistin concentration using a liquid chromatography-tandem mass spectrometry (LC-MS/MS) system. The population PK model was developed using nonlinear mixed effect modeling (NONMEM, v. 6.2). A one-compartment linear PK model for colistin best described the data. The covariates included in the final model were creatinine clearance for the relative fraction of CMS converted into colistin and the presence of edema for the turnover rate constant of CMS converted into colistin. A steady-state 24-h area under the concentration-time curve was simulated from 1,000 virtual patients receiving 150 mg colistin base activity every 12 h using the final model. Relative to previous studies with critically ill patients, the elimination half-life of colistin (6.6 h) was much shorter, and continuous renal replacement therapy was not a significant covariate for any PK parameters.