There is increasing interest in renewable energy sources interconnections in distribution systems since they are inexhaustible and nonpolluting. Wind and photovoltaic are among most mature renewable energy sources, and their penetration continues to increase. This paper proposes a method for distribution network reconfiguration to minimize annual energy losses by determining the optimal configuration for each season of the year. Uncertainties including the daily time varying load and stochastic power generation of renewable distributed generators (DGs) are modeled and taken into account. The method is based on generating a probabilistic generation-load model that combines all possible operating conditions of the renewable DGs with their probabilities of occurrence, then accommodating this model in the reconfiguration problem. The reconfiguration problem, based on Genetic Algorithm, aims at achieving minimum annual energy losses. The constraints include the voltage limits, line current limits, radial topology, and feeding of all loads. The proposed method has been tested on two systems. Simulation results show significant reductions in the seasonal and annual energy losses for all the studied cases.