A brain–computer interface decodes in near-real time the imagined speech of people who have difficulty enunciating words.