In this paper an elliptical reflectarray antenna is optimized by some novel evolutionary techniques: Ant Colony Optimization (ACO), Biogeography Based Optimization (BBO), Differential Evolution (DE), Population-Based Incremental Learning (PBIL) and Stud Genetic Algorithm (SGA). The optimized design of the reflectarray is aimed to compare their performances on a complex EM optimization problem. Results show that some techniques (DE, BBO, SGA) are particularly effective in dealing with antenna optimization