This study tests the applicability of Linear Discriminant Analysis for discriminating between speakers based on the trajectories of vowel formants, using real-case forensic recordings.