Unbound daptomycin concentrations are responsible for pharmacologically beneficial and adverse effects, although most previous reports have been limited to the use of total concentrations. We developed a population pharmacokinetic model to predict both total and unbound daptomycin concentrations. Clinical data were collected from 58 patients with methicillin-resistant Staphylococcusaureus including patients undergoing hemodialysis. A total of 339 serum total and 329 unbound daptomycin concentrations were used for model construction. Total and unbound daptomycin concentration was explained by a model that assumed first-order distribution with two compartments, and first-order elimination. Normal fat body mass was identified as covariates. Renal function was incorporated as a linear function of renal clearance and independent non-renal clearance. The unbound fraction was estimated to be 0.066 with a standard albumin of 45g/L and standard creatinine clearance of 100mL/min. Simulated unbound daptomycin concentration was compared with minimum inhibitory concentration as a measure of clinical effectiveness and exposure-level-related induction of creatine phosphokinase elevation. The recommended doses were 4mg/kg for patients with severe renal function [creatinine clearance (CLcr) ≤30mL/min] and 6mg/kg for patients with mild to moderate renal function (CLcr>30 and ≤60mL/min). A simulation indicated that dose adjusted by body weight and renal function improved target attainment. This population pharmacokinetics model for unbound daptomycin could help clinicians to select the appropriate dose regimen for patients undergoing daptomycin treatment and reduce associated adverse effects.