This study proposes a Brain–Computer Interface (BCI) based on the Steady-State Visual Evoked Potential (SSVEP). This BCI can discriminate one out of four classes, once a second. Using such BCI, nine healthy volunteers were able to use the four classes with an average precision of 83 15�%. Moreover, three of such volunteers were selected to guide a robotic wheelchair through an indoor environment using such BCI, and the result is that all of them were able to accomplish the proposed task. For that, four flickering visual stimuli were used and each one was associated to one of the BCI classes. Once a stimulus was observed, a command associated to it was identified and translated to a visual feedback or to a wheelchair command. The electroencephalogram signal of the volunteers was acquired and processed according to a feature extraction and a classification steps, leading to the identification of the stimulus the user gazed.