A number of recent studies have successfully demonstrated smart appliances control using brain-computer interface (BCI) as a proof-of-concept. Yet, a fluid and intuitive interface for the actual implementation is still nowhere to be found. As a first step in bringing a BCI system for smart TV control to a reality, this study focuses on isolating factors that should be considered in design. In the experiment, the participants selected a smart TV channel with different distances to the TV set using a noninvasive BCI that detected the P300 waveform from EEG. The channel menus were presented as either a still image or a video. Various viewing distances as well as effects of video in the stimuli are tested. While video menu presentation caused a decrease in performance as expected, there were unexpected influences from viewing distance as the closest distance did not yield the highest accuracy. Our results suggest a putative optimal distance for P300-based BCI control of smart TV, which was also consistent with that from previous ergonomics studies and several more BCI design components to be experimented on.