Taking an area of Xi’an city as an example, the Storm Water Management Model was established according to the measured data after calibration and validation. Three different return period storms (once in 2 years, once in a decade, and once in 20 years) were designed to simulate the water quantity and quality of the study area with or without rain gardens. Rain gardens with different proportions (1, 2, and 4 %) of the study area were set for simulation and comparison. Finally, the control effects of rain gardens with and without discharge facility were compared. Rain gardens can reduce the total amount and peak flow of urban runoff. Such gardens can also reduce pollutant peak concentration and pollution load. The peak time of flow and concentration can be delayed. However, the control effect or reduction rate of rain gardens will decrease with increasing rainfall intensity. A larger area proportion of rain gardens results in a better control effect on urban runoff. The control effect of rain gardens without discharge facility is better than that of rain gardens with discharge facility under the same conditions of rainfall and area ratio. The control effect difference between rain gardens with and without discharge facility decreases with the increase in the area ratio of rain gardens.