One of the main intent of smart grid is to enable more renewable energy resources. Installing the DGs near endcustomers result in reduction of power transmission of lines, however, if a large amount of microDGs generate energy simultaneously, it may result in distribution system congestion. In usual, distribution component capacity is considered high enough to avoid any congestion. However, considering the on-line monitoring and controlling abilities of a smart grid, this paper addressed reconfiguration as an effective methodology to solve the distribution network congestion problem. Genetic algorithm is used to determine the optimal configuration. Numerical results of implementing the method on a test system demonstrate that network reconfiguration can be used as an effective methodology to defer the additional investment in distribution systems.