Kinetic power, i.e. kinetic energy per unit time and area, is the variable widely used to represent the rainfall erosivity which affects soil loss and sediment yield. This paper shows the results of an experimental investigation using the raindrop size distributions (DSDs) measured by an optical disdrometer installed at the Department of Agricultural, Food and Forestry Sciences of University of Palermo in Italy (June 2006-March 2014) and at the El Teularet experimental station in Spain (July 2015-May 2016). At first an analysis of the DSDs aggregated into intensity classes is carried out, then the measured kinetic power values are determined. The aggregated DSDs allowed to establish that the median volume diameter of the distribution is affected by raindrops characterized by the greatest values of the diameters that composes precipitation. The measured kinetic power values allowed to verify the reliability of kinetic power-rainfall intensity relationships proposed by Wischmeier and Smith and Kinnell. Finally, using all the available measurements of kinetic power, rainfall intensity and median volume diameter obtained in different climatic contexts and by different measurement techniques, this paper demonstrates that the ratio between kinetic power and rainfall intensity depends strictly only on median volume diameter of the distribution according to a single site-independent relationship. Therefore the estimate of the kinetic energy per unit volume of rainfall does not require the knowledge of the whole drop size distribution. The reliability of a theoretical relationship relating the kinetic power per unit volume of rainfall to median volume diameter is also positively verified using all available measurements.