Urban growth in watershed basins has left many negative impacts on water resources throughout the world and has reduced permeability. Increasing the area of impermeable surfaces is influential on both runoff characteristics, quantity and quality. Several studies have shown that urban runoff contamination is comparable with sewage pollution. As a result, modern runoff management should simultaneously manage flood and pollution. In this research, source control methods are used to control quantitative and qualitative flood. Best Management Practices or BMPs are referred to as a set of structural and non-structural measures that are used to improve the quality and quantity of runoff. Three types of BMPs are discussed in this study. Bioretention systems, wetlands and Infiltration systems. Each of these BMPs has different requirements for installation, cost and reduction efficiency. The purpose of this research is to find the least costly combination of these three types of BMPs that can satisfy the EPA quality standards of surface water in Sepahan-Shahr area. The MUSIC model was used to obtain the ability of BMPs to remove the pollutants. This elimination was calculated as the percentage of removal of various pollutants by BMP and also the concentration of each pollutant at the outlet point. It is necessary to model the number of combinations from a BMP to be able to omit the collective of them. Considering that no quantitative and qualitative characteristics of the urban flood in the study area have been systematically recorded, design and selection of the best combinations of BMPs was undertaken based on three suggested concentrations in urban basins by the US Environmental Protection Agency (EPA). The LINGO model was used to find the optimal answer. The outputs of the quality model were entered as inputs to the LINGO optimizer and with the lifetime expense of each BMPs, the least costly combination was selected in each of the three scenarios mentioned. Sensitivity analysis was done on some variables, which may affect final results. The results show that when the amount of pollutants is at its lowest level, the construction of 3 biological systems with a total area of 600 square meters is the least costly option. Also, a combination of four biological systems and one infiltration system with a total area of 900 square meters is selected for the time when the pollutants have a moderate concentration. When pollutants have the highest observed concentration in urban areas, 10 biological systems with a total area of 2000 square meters are selected as the most optimal combination. The overall results showed that bioretention systems are the best method for treatment of water in the study area. This is due to the relatively low cost of bioretention systems compared to other BMPs and the high ability to remove pollutants. Also, the results of sensitivity analysis showed that changing the permeability and the quality standards affect the final results of modeling and the result of the model is sensitive to these parameters. But considering the renewal and decommissioning cost, life cycle change, as well as changing the inflation rate and interest do not affect the final results of the model. Key Words Urban runoff, BMP, Contaminant, Modelling, Bioretention system, MUSIC,LINGO, Life Cycle Cost.