Recently, there has been rapid development in the field of proton exchange membrane fuel cell (PEMFC) systems for transportation applications. The performance of a PEMFC system is very sensitive to the operating conditions, and uncontrolled working conditions may cause malfunctions and degradation. A robust control strategy is urgently needed, in order to improve the reliability of PEMFCs and prolong their working lifetime. To develop such a control strategy, one needs to not only model the system with identified parameters, but also know their uncertainties. In most studies related to system uncertainties, however, the parameter uncertainty is usually regarded as a pre-known condition. This paper proposes a method to identify key parameters and their boundaries of PEMFCs, and analyze the uncertainties of internal states based on Monte Carlo simulation. A nonlinear isothermal dynamic model, which takes into account the filling and-emptying dynamic sub-models and a sub-model of mass transport through the membrane, is firstly introduced. Key parameters are then extracted stepwise using a nonlinear least squares (NLS) algorithm, and the parameter boundaries are identified based on Monte Carlo simulations. The uncertainties of internal states in time and frequency domains are investigated afterwards. The results demonstrate the effectiveness of this method. Among the three sub-systems (cathode, anode, and membrane), the cathode sub-system was found to have the smallest uncertainties, while the membrane has the largest. Transfer functions for small disturbances of cell current to the internal states also have uncertainties, which can be low-frequency pass or bandpass functions depending on the parameter values. Further study will focus on the design of robust control strategies based on system models with uncertainties. (C) 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd, All rights reserved.