Though, traditionally, electrophysiological recordings have been limited to provide temporal information on neural activity, the development of mathematical algorithms capable of solving the inverse problem is facilitating, in recent years, the access to spatial information (i.e., on the origin of neural activation). This study explored a new strategy in order to increase the reliability of inverse problem solutions: applying these algorithms on factor scores (and not on voltages), a parameter that can be defined as "clean amplitude". Factor scores derive from Principal Component Analysis (PCA) applied to event-related potentials (ERPs). The main advantage of PCA is its capability to extract and quantify ERP components free of the influence of adjacent or subjacent components. The LORETA algorithm for source localization was applied on peak voltage, average voltage and factor scores for the motor potential recorded from 25 subjects, who had to repeatedly press a button with their right hand. The solutions given by LORETA in these three modalities were compared. The motor potential, a negative wave that begins just before any voluntary movement and is centrally distributed in the scalp, is particularly useful to the scope of this study, since its origin is known: contralateral motor cortex. Results show that the three modalities (peak voltage, mean voltage and factor scores) provided the same main focus (left motor cortex), though the "cleanest" solution (i.e., the main focus was more salient with respect to other secondary, noisy foci) was achieved by the factor score-based LORETA.