In order to study the sensitivity of water quality parameters of different underlying surface types under various rainfall scenarios, the modified Morris screening method was used to analyze the overall characteristics of water quality parameters sensitivity of the SWMM, clarify the influence degree of different rainfall intensity and underlying surface types on water quality parameter sensitivity, and reveal the differentiation characteristics of highly sensitive parameters of each water quality index. The results show that the number of highly sensitive and sensitive parameters and sensitivity increase with increasing rainfall intensity, and the sensitivity level of water quality parameters for road and grassy area significantly elevated under rainstorm condition, however, the time lag of rainfall peak will reduce the sensitivity of some highly sensitive water quality parameters. In terms of sensitivity, the parameters are runoff index, maximum accumulation, washoff coefficient, rate constant and half-saturation constant in descending order. The significance tests show that the influence degree of underlying surface types on the number of highly sensitive and sensitive parameters is higher than that of rainfall conditions. The number of sensitivity parameters and their sensitivity for roof are generally large, mainly affected by the area proportion and permeability of underlying surfaces. The sensitivities of runoff exponents of ammonia nitrogen for farmland are significantly higher than COD, TP and TN, which is related to the process of fertilizer hydrolysis and ammonia nitrogen loss.