In this article, a novel method is proposed to help evaluating the reliability of a Proton Exchange Membrane Fuel Cell (PEMFC) stack. The method is based on a knowledge fuel cell model which is one-dimensional, non-isothermal and includes a diphasic representation of the fluidic flows (each reactant is considered as a mix of gases and liquid water). The modeling is implemented using the DYMOLA software developed by Dassault Systèmes and based on the MODELICA language. One of the advantages of this simulation tool is that it allows a close coupling of multi-physical modeling with statistical treatments. In this perspective, the modeling is done with the aim of having as much relevant physical parameters as possible. The different effects of these parameters on the PEMFC electrical behavior can be observed and the performance sensitivity can be determined considering statistical distributions of input parameters, which is a step towards reliability analysis.