This thesis deals with a stochastic version of the simple facility location problem where the demands of customers are random variables with given density functions, in contrast to the traditional location model which assume fixed demands. The model considered in this thesis is approximated to the equivalent mixed integer linear programming problem(MILP) which can be viewed as a special case of ordinary deterministic location problem. Then the powerful dual ascent method by Bilde \& Krarup[2] is shown to be applied to this MILP with only minor modifications. Computational experiments with 16 problems are provided.