This paper presents a reliability approach to analyze the degradation of proton exchange membrane fuel cell. This approach is based on the dependability analysis tools such as the causal and fault trees to establish an analysis of the internal state of the fuel cell energy conversion performance and evaluate its lifetime. The elaboration of causal tree offers powerful tools to a deductive analysis, which consists on seeking the various combinations of events leading to the fuel cell degradation. The parameters of fuel cell model are identified in order to found the degree of degradation. The experimental determination of the variation interval of the parameters is done according to each of degradation modes. A diagnostic method is proposed in order to identify the depth of each aging process of the fuel cell. The diagnosis is done by comparing the experimental output characteristic at beginning of life of the fuel cell with the used fuel cell to qualify and quantify the depth of degradation.