Decision making on budgeting is one of the most important issues for executing managers. Budgeting is a major tool for planning andcontrol of projects. In public and non-profit organizations and institutions, estimating the costs and revenues plays an important role inreceiving credit and budgeting. In this regard, in the present paper the case of Isfahan municipality is considered. One of the mainexpenditures of the 14 districts of Isfahan is the costs related to water. Predicting the total cost of water helps the municipality of Isfahan tooptimize the water use in its 14 urban zones. Thus, in this study the total cost of water in the districts of Isfahan is estimated usingregression analysis and neural network models. Then the results of the methods are compared with each other to minimize the deviationsfrom the approved budget. Finally, the neural network method is selected as the main simulation method for forecasting the total cost ofwater in the districts of Isfahan. Manuscript profile