The State of charge (SoC) of a lead-acid battery is linearly proportional to its open circuit voltage, which cannot be measured because the battery is continuously connected to the system that integrates it. This paper proposes an approach for estimating the open circuit voltage as a function of parameters of the electrical equivalent model. This approach is based on the recursive algorithm of Kalman Filter to estimate the State of charge of a battery dynamically. We also developed a state representation of a battery. Simulation and experimental validation of the model have been applied to a 'SLT210' lead-acid battery connected to a photovoltaic panel. Good concordance between estimation and measurement results is shown over periodic discharge, random discharge and charge/discharge cycles. (C) 2012 Elsevier Ltd. All rights reserved.