This paper is in the field of electric power quality monitoring and presents a new approach for the identification, classification and characterization of the nine voltage dips and swells in electricity networks. The proposed method is based on the study in the complex plane of the signatures of the different voltage dips and swells. In the study of these signatures, the elements taken into account are the root mean square (RMS) values of the phases, the existence or not of an additional phase shift of the voltages and their rotation sense. The informations obtained are synthesized in three variables, and used to the implementation of the method. The results found by the computer simulations carried out by the MATLAB/Simulink software show that the proposed approach uses few parameters, is easy to implement and understand, and makes it possible to efficiently detect, classify and characterize the nine voltage dips and swells.