Simulation of a proton exchange membrane fuel cell (PEMFC) in its assembled state has been a long-standing challenge, with one factor being the influence of gas diffusion layer compression resulting from heat and mass transfer on the effective proton conductivity of the proton exchange membrane. Due to the lack of in-situ data, it is customary to utilize an empirical formula as a conventional model for determining the effective proton conductivity. However, significant deviations (>10%) have been observed between simulated and experimental data for fuel cells, mainly when the fuel cell is assembled. The assembly of PEMFC caused a shift of effective proton conductivity, leading to a significant deviation. To address this issue, this study proposes a model using COMSOL that integrates mechanics, electrochemistry, heat, and mass transfer of the fuel cell. To decrease the deviation between simulation and experiment, the effective proton conductivity of the proposed model is corrected by the reference proton conductivity. Specifically, an adjustment factor is introduced to the reference proton conductivity to correct the shift of effective proton conductivity caused by the compression. As a result, the average deviation of the proposed model is decreased from 10.44% to 2.25%, compared to a traditional model. As a case study, the optimal compression ratio of 20% is obtained by heat and mass transfer analysis, in which peak power density is increased from 6611.2 to 7466.6 W m−2. This study highlights the importance of membrane proton conductivity for the output performance of PEMFC.