Power quality monitoring is assuming an increasing importance because of the large diffusion of nonlinear and time-varying single-phase and three-phase loads and their interactions with the power supply system. This has made that the industrial electrical consumer systematically uses measure equipment to evaluate the electric power quality. The measurements show that the harmonics of the waveforms are variable along with the time due to the changes in the load conditions and in the system configuration. Thus, the electrical users dispose of a great quantity of data from such of the three phases of the supply. Therefore, it is necessary to find a way to characterize this great group of registered data. Thus, the objectives of this work are: firstly, to propose two indices, a three-phase total demand distortion index and a three-phase total unbalance index; secondly, this paper proposes a simple way to characterize the groups of data in an statistical way which make easy its representation. The study carried out is applied to the data recently registered in an electrical power system.