EEG based brain computer interface has emerged as a hot spot in the study of neuroscience, machine learning and rehabilitation in the recent years. A BCI provides a platform for direct communication between a human brain and a computer without the normal neurophysiology pathways. The electrical signals in the brain, because of their fast response to cognitive processes are most suitable as non-motor controlled mediation between the human and a computer. It can serve as a communication and control channel for different applications. Though the primary goal is to restore communication in severely paralyzed population, the BCI for speech communication fetches recognition in a variety of non-medical fields, the silent speech communication, cognitive biometrics and synthetic telepathy to name a few. A survey of diverse applications and principles of the BCI technology used for speech communication is discussed in this chapter. An ample evidence of speech communication used by "locked-in" patients is specified. Through the aid of assistive computer technology, they were able to pen their memoir. The current state-of-the-art techniques and control signals used for speech communication is described in brief. Possible future research directions are discussed. A comparison of indirect and direct methods of BCI speech production is shown. The direct method involves capturing the brain signals of the intended speech or speech imagery, processes the signals to predict the speech and synthesizes the speech