Introduction: Essential oils (EOs) obtained from medicinal plants are complex mixtures of chemicals with wide spectrum of biological activities. However, the use of these oils in pharmaceutical formulations is limited due to their lipophilicity, high susceptibility to degradation and volatilization. In this context, encapsulation of EOs in polymeric nanoparticles is presented as a flexible and efficient approach to improving their physical stability, aqueous solubility and therapeutic efficacy1. In the preparation of nanosystems, the statistical tools help in problems evaluation and process parameters optimization2. The aim of the present work was to evaluate the impact of variables related to the preparation process on polymeric nanoparticles containing Cymbopogon citratus essential oil (DC.) Stapf (CcEO). Methodology: PLGA nanoparticles were prepared by emulsification-solvent diffusion method3. In order to optimize this process, Box-Behnken design was developed using Statistica v. 6.0, which generated a project matrix of 17 experiments.The influence of the variables such as sonication time (X1), amplitude of ultrasound power (X2) and ratio CcEO/PLGA (X3) on the response encapsulation efficiency (Y1) was evaluated. Y1 was determined by solvent extraction. Results: In the determination of the factors that affect Y1 it was verified that particles with higher content of encapsulated CcEO are prepared by increasing X2 and reducing X3, whereas X1 presented no influence on this response. In the optimization process, particles with encapsulation efficiency of 76.70% were obtained using sonication time of 8 minutes, amplitude of ultrasound power of 150 W and ratio CcEO/polymer of 0.5. Conclusion: The results suggest the influence of the amplitude of the ultrasound power and ratio CcEO/polymer variables in the efficiency of encapsulation of the nanoparticles used with delivery systems for CcEO.