The aim of this paper is to present a classification method that is capable to discriminate between Event Related Potentials (ERPs) that are the result of observation of correct and incorrect actions. ERP data from 47 electrodes were acquired from eight volunteers (observers), who observed correct or incorrect responses of subjects (actors) performing a special designed task. A number of histogram-related features were calculated from each ERP recording and the most significant ones were selected using a statistical ranking criterion. The Support Vector Machines algorithm combined with the leave-one-out technique was used for the classification task. The proposed approach discriminated the two classes (observation of correct and incorrect actions) with accuracy 100%. The proposed ERP-signal classification method provides a promising tool to study observational-learning mechanisms in joint-action research and may foster the future development of systems capable of automatically detecting erroneous actions in human-human and human-artificial agent interactions.