Resilience intervention is used to train and improve one's resilience level, however current studies used various psychometric instruments for resilience evaluation that are subjective, tedious, and prone to bias. In this study, electroencephalogram (EEG) has been chosen as a potential candidate for a new resilience assessment method that aims to rectify the limitations of existing psychometric instruments. The aim of the work is to create a practical resilience assessment system using EEG, which can classify low or high resilience level with high accuracy and generalization. The proposed system uses Virtual Reality headset to simulate rest and task conditions during EEG data acquisition. Connor-Davidson Resilience Scale (CD-RISC) is used to determine the actual resilience score of each participant. Results indicate that Linear Discriminant Analysis (LDA), Naïve Bayes (NB), and Support Vector Machine (SVM) using ReliefF for feature selection yields the best model performances, however other feature extraction and selection methods need to be explored to improve predictive strength of features on resilience. © 2024 IEEE.