Brain computer interface systems use brain signals to enable the control of external devices, such as: wheelchairs, communicators, neuro-prosthesis, among others; in people with severe motor disabilities. In this study two young men with motor disabilities were trained to learn how to control a brain computer interface (BCI) using the P300 evoked potential recorded by electroencephalography (EEG). This interface enables the verbal communication through a stimulation matrix (4x3 rows and columns) that contains the picture of current football players and a sound related to each one which reproduced the name of them. Initially, we did the preparation stage that allowed the users to learn how to control the BCI. After this, the real probes were performed in two stages: the training tests and the online tests, where the participant could select the icons that he wanted through his brain signals. The average percentages of success for each participant were: 95% and 85%, respectively. The average time for an icon selection was 36 seconds (i.e.1.67 letters/minute).