In this study, we created 8-command P300 tactile brain-computer interface, running on minimally modified consumer Braille display, and tested it on 10 blind subjects and 10 sighted controls with two stimuli types, differing in size. Larger stimuli provide better BCI performance both in blind and sighted participants than smaller stimuli. With large stimuli, median target selection accuracy in the blind group was 95%, which is 27% more than sighted controls (p < 0.05), suggesting that blind subjects are not only able to use tactile brain-computer interface but also can achieve superior results in comparison with sighted subjects. The difference in event-related potentials between groups is located in frontocentral sites around 300 ms post-stimulus and corresponds with early cognitive event-related potential components. Blind subjects have higher amplitude and shorter latency of ERPs. This effect was consistent across stimuli types. This is the first study to evaluate differences in event-related potentials between blind and sighted subjects in a BCI-specific task.