In this study, the drying process of Brazilian pepper-tree fruits (Schinus terebinthifolius Raddi) was investigated, with emphasis on the development of models for the drying kinetics. A central composite rotable design (CCRD) with 11 experiments was carried out. In each experiment, the drying kinetics and the drying rate were observed. Classical models were fitted to the drying data, as well as a model based on Artificial Intelligence. The results revealed that the air temperature has the greatest influence on the process in the range of the experimental conditions. The modeling by artificial neural network and the Henderson model showed the most suitable results.