In today’s sophisticated networks, installation of energy storage systems (ESS) is felt completely essential due to intermittent energy generation from renewable resource stochastic behavior of loads and reliability issues. Current work addresses a new method based on stochastic unit commitment to size the battery energy storage systems (BESSs) optimally in microgrids (MGs) where plug-in electric vehicles (PEVs) will be used. Stochastic variables such as electric loads and wind speed are modeled with the corresponding probability density functionsPDFs). The optimization problem is then formulated as a stochastic mixed integer linear programming in GAMS environment. Numerical simulation and results are provided for a number of case studies to expose the model capabilities