The proposed research aims to derive individual neural characteristics from data-driven analysis of resting state fMRI to predict the effectiveness of Mindfulness-Based Stress Reduction (MBSR) training on fear extinction and stress reduction. This project aims to capture the individual variability in the amplitudes of spontaneous brain activity and neural network properties, and relate such variability with individual behavioral traits of anxiety, anger and emotion regulation capacity, and use such variability to predict the amount of effect of MBSR on each individual. The variability of the effect of MBSR will be characterized with changes in the scores of the Perceived Stress Scale, the performance of the fear conditioning task, and the magnitude of neural response in the fear conditioning task. Structural Equation Modeling and machine learning techniques will be used to evaluate the prediction power of different neural characteristics. This project will address the important question of individual variability in contemplative neuroscience research and evaluate the variability of neural characteristics and behavioral effectiveness quantitatively.