In the distribution network, with the permeability of distributed generator increasing, the influence on grid voltage and the network loss will become apparent. This effect is closely related to the location and capacity of distributed generator. A method of planning the timing distributed generator is proposed. First installation node of DG is selected by the comprehensive sensitivity analysis of node voltage and net loss, then optimization methods are used for distributed power capacity optimization. The optimization method is, establishing the objective function in order to reduce the network loss and improve the level of voltage, and establishing the conditions by chance constrained programming method; and doing random flow calculation by using the Monte-Carlo simulation method for random scheduling of DG and load; finally, the genetic algorithm of elitist strategy is employed to solve the model. Results verify the effectiveness of the method of site selection and capacity.