Micro-expression recognition is an active research domain due to promising applications ranges from lie detection to reveals genuine emotions. Micro-expression recognition is a challenging task due to repressed facial appearance and subtle nature of emotion. This paper investigates implementation of curvelet transform, a multi-directional and multiresolution algorithm for micro-expression recognition. Discrete Curvelet Transform has proven its importance in various image processing applications. However, we are using first time (to the best of our knowledge) for micro-expression recognition. The curvelet algorithms are implemented over CASME-II, a benchmark database for micro-expression recognition. The experimental results reveals that curvelet based method provides best accuracy compare to existing approaches reported in literature.