This paper discusses a novel method for optimum sizing battery storage capacity for distributed power system utilizing chance constrained programming. It first develops a detailed chance constrained model of sizing batteries, which deal with the stochastic characteristics of wind power, photovoltaic power and loads in distributed power system expressed by means of probability. Genetic algorithm is used to solve the problem. The results on sizing battery demonstrate the advantages of the proposed approach. Keywords?distributed power system, sizing battery storage capacity, stochastic simulation, genetic algorithm.