Aim: To establish a population pharmacokinetics (PPK) model for lamotrigine (LTG) in Chinese children with epilepsy in order to formulate an individualized dosage guideline. Methods: LTG steady-state plasma concentration data from therapeutic drug monitoring (TDM) were collected retrospectively from 284 patients, with a total of 404 plasma drug concentrations. LTG concentrations were determined using a HPLC method. The patients were divided into 2 groups: PPK model group (n=116) and PPK valid group (n=168). A PPK model of LTG was established with NONMEM based on the data from PPK model group according to a one-compartment model with first order absorption and elimination. To validate the basic and final model, the plasma drug concentrations of the patients in PPK model group and PPK valid group were predicted by the two models. Results: The final regression model for LTG was as follows: CL (L/h)=1.01(star)(TBW/27.87)(0.635 star)e(-0.753 star VPA star)e(0.868 star CBZ star)e(0.633 star PB), Vd (L)= 16.7(star)(TBW/27.87). The final PPK model was demonstrated to be stable and effective in the prediction of serum LTG concentrations by an internal and external approach validation. Conclusion: A PPK model of LTG in Chinese children with epilepsy was successfully established with NONMEM. LTG concentrations can be predicted accurately by this model. The model may be very useful for establishing initial LTG dosage guidelines.