The Soil Conservation Service Curve Number Model, proposed by the U.S. Department of Agriculture (USDA), has only one parameter CN, and is a tool for predicting runoff. According to the SCS-CN methodology, soils are categorized into four distinct hydrologic soil groups (HSGs) based on their inherent ability to generate runoff. However, the delineation of these four discrete HSG levels can lead to abrupt shifts in the Curve Number (CN) value as one category transitions to another. To obtain more accurate CN values that better reflect the hydrological soil conditions in China, CN values for each HSG were assessed using both the median method (CN_M) and the least squares fit method (CN_F) based on monitored rainfall-runoff data from 48 sites across China. These values were found to significantly deviate from the curve number values (CN_T) provided in the USDA-SCS handbook. The findings indicated that replacing CN_T with CN_F, derived through the least squares fit method, improved the efficacy of the conventional SCS-CN approach. Nevertheless, CN_F exhibited suboptimal performance within HSGs A and B. The subpar performance could be attributed to the significant variability in CN values observed within each hydrological soil group. Therefore, the proposed model taking the influence of soil saturated hydraulic conductivity (Ks) on runoff prediction into account was developed to reflect the influence of CN changes under different soil types. The proposed method underwent a reliability test using data from 44 study sites, and subsequently, it was carried over into the remaining 4 typical sites, employing parameters calibrated using the initial 44 sites data. The proposed method with high NSE and low RMSE values demonstrated remarkable predictive precision for runoff at the sites, surpassing the original SCS-CN approach regardless of using CN_F or CN_T. Hence, the proposed method offers versatility and is advantageous for widespread use across China's diverse hydrological soil environments.