Nowadays, urbanization is a common process which replaces vegetation cover with impervious areas. This process increases urban stormwater. A new tendency in urban stormwater management endorses ‘source control’, whereby distributed water sensitive urban design systems are built throughout the subdivisions to alleviate the effects of land use changes. Various costs and functions are considered for different urban stormwater treatment measures methods. The present study introduced a legal optimization approach, to minimize the costs of urban stormwater treatment measures. For this purpose, the MUSIC model and Genetic Algorithm were combined in the Matlab environment. The Results of application of MUSIC-GA model, to optimize urban stormwater treatment systems, at 2.8 hectares of industrial areas near Rowzehchay River in Lake Urmia basin, showed that the developed MUSIC-GA model has an efficient performance for finding the optimal urban stormwater control. The results also indicated that the optimized treatment measure in the post development scenario decreased at least 45 percent of pollutants from urban stormwater runoff. Moreover, very small values of coefficient of variation (0.00007) among different results of multiple runs indicated that there was a high convergence between result of MUSIC-GA and the global optimal solution.