To enhance the efficiency of continuous cement mixing in well cementing, a Computational Fluid Dynamics (CFD) approach was employed to investigate the mud flow and mixing processes within the continuous mixing system. This study aims to analyze the influence of various design factors and their interactions on the mixing of different slurries in the well cementing slurry continuous mixer. Using the Response Surface Methodology (RSM) with a Box-Behnken model, this work combines the Euler-Euler multiphase flow model, k-turbulence model, and the Multiple Reference Frame (MRF) method to simulate two-phase fluid flow and agitator rotation, establishing the CFD model of the mixing system. The CFD model, coupled with the RSM method, analyzed four design factors: impeller diameter (400–800 mm), design density (1.4–2.6 g/cm³), impeller speed (100–300 rpm), and particle size (20–500 μm), studying their impact on two response variables, slurry mixing uniformity, and total power consumption. Through RSM and multiple regression analyses, predictive models for slurry mixing uniformity and corresponding power consumption were established under different design factors. Data comparison analysis revealed excellent agreement between predicted and actual results. Findings indicate that, when using uniformity or power as evaluation criteria, the P-value is smallest for impeller speed and largest for particle size. Hence, the impeller speed exhibited the strongest correlation with both, followed by impeller diameter, and lastly, design density and particle size. Only the interaction between impeller diameter and impeller speed (P-value <0.0001) significantly influenced both power and mixing uniformity, with its significance ranking just below the linear effects of impeller speed and diameter. Slurry uniformity decreased with increased design density and particle size, while power consumption exhibited the opposite trend. Uniformity increased with increasing impeller speed, but when the uniformity reached its maximum, further speed increments led to a slight decrease in uniformity, accompanied by a sharp increase in power. This was significantly influenced by the quadratic effect of impeller speed (P-value <0.0001), with its significance weaker than the linear effects. The proposed predictive model accurately forecasts response values for any given design parameter, providing a research direction for optimizing mixer design and selecting corresponding design parameters. This approach streamlines the study of mud mixing systems, reducing research time and costs. Moreover, it offers scientific guidance for optimizing operational parameters corresponding to well cementing conditions.