Abstarct: It has been forecasted that by 2025, Iran will face with water crisis. It is predicted that available water per capita would fall between 776 to 860 cms in 2025. Water production is not the only solution to overcome the problem of water shortage, but the country must prevent waste and losses of water with proper management and optimal utilization of prevailing resources, to increase the efficiency and lifetime of water systems and minimize costs. Obviously, the optimal management of water systems and decision-making requires information and tools to monitor available water resources. So, an optimal monitoring network, for estimating and predicting of the groundwater level is an essential tool for aforementioned aim. The objective of this study is to find the best sites for monitoring wells at Tehran_Karaj aquifer. To find the optimum location for wells, a combination of the Kriging interpolation method and genetic algorithm was applied. The objective function was to minimize the groundwater level estimation error i.e. the best solution to the problem is the coordinates of the Kriging implementation that have lowest total estimation error. Results revealed that by shifting the coordinates of 57 wells out of 164, estimation error would decrease to a considerable amount, i.e. one third of the original error.