Polymer electrolyte membrane fuel cells are proving to be a clean and efficient source of energy. Nowadays, extensive research efforts are being focused on bringing this technology to everyday use. An important aspect when integrating fuel cells in practical applications is their ability to respond to load demand. With respect to this, due to their complex internal dynamics, fuel cells belong to the group of more slowly responding sources. In order to make them more generally applicative they are often connected with a battery or a super-capacitor via a power converter to form a hybrid power source. A control algorithm, designed for such a system, represents an interesting challenge: it has to adapt to varying working conditions and operate optimally in terms of efficiency and reliability, while minimizing any impacts on the degradation of the components. Here, we present an approach using supervisory control automaton that switches between the system's operational modes and sets the references for the lower-level control loops. The evaluation of the efficiency and degradation is carried out in a simulation using a model of the widely used 1.2-kW Ballard Nexa power module. (C) 2011 Elsevier B.V. All rights reserved.