The objective of the present study was the formulation of the theoretical premises on which to base the approaches to the overcoming of the limitations not infrequently imposed on the sound records of the speech material of interest. The secondary objective was the development of the methods for this purpose. We analyzed the literature sources, the methodological instructive regulations, and the results of medical criminalistics investigations. The study has demonstrated that the spectral analysis of the speech materials provides a reliable tool for overcoming the incomplete (by force of different causes) compatibility between the initial phonograms and those under consideration for the purpose of forensic medical expertise with a view to the identification of personality of the speaker.