Abstract: As a system to communicate between patients with communication disorders or physically disabilities and machines, brain-computer interface technologies have received widespread attentions in recent years. Meanwhile, as a new type of brain-computer interface, brain-computer interface for speech imagery develops rapidly because of its potential to offer an effective and comfortable communication tool to patients suffering communication disorders. This review paper first introduces common signal acquisition technologies used in existing brain-computer interfaces for speech imagery, and presents in details the signal processing algorithms in speech-imagery-targeted brain-computer interface studies. Finally, some practical considerations and discussions are provided on the existing research problems and future developments of brain-computer interface technologies for speech imagery.