<p>Pressure differential deviations under static conditions and pressure convergence fluctuations under dynamic disturbances are widely reported problems with pressure differential control in pharmaceutical cleanrooms, yet their underlying mechanisms and key reasons remain insufficiently explored. This study performed a field survey and model-based simulations to identify the major influencing parameters and quantify their influence on pressure differentials. Twelve pharmaceutical cleanrooms with varying environmental control parameters were included in the field survey, all of which were served by a variable air volume (VAV) ventilation system. Large deviations between actual and design pressure differentials were found, ranging from 10% to 42.5%, and a total of 24 uncertain parameters and their respective uncertainty ranges were identified. Based on the field survey, a data-driven pressure differential response model was developed using MATLAB/Simulink platform. The model fully took into account the system dynamics and facilitated real-time monitoring and control of the pressure differential. Sobol-based sensitivity analysis was then conducted to identify key influencing parameters of pressure differential deviations. The simulated results revealed that static pressure differential deviations were predominantly influenced by pressure sensing accuracy, exhaust airflow accuracy, and duct impedance, while dynamic disturbances were mainly driven by room envelope airtightness and supply airflow accuracy. The interactions between connected zones were pronounced. Rooms with higher branch duct impedance experienced smaller pressure differential deviations due to natural buffering characteristics, while the parameter uncertainties in these rooms significantly affected pressure differential in other rooms. These findings offer practical guidance for the design and operation of precise pressure differential control in pharmaceutical cleanrooms.</p>