This book is devoted to the theories and models underlying the design of current text-to-speech synthesizers. It was inspired by presentations at the ESCA Tutorial Day & Workshop on Speech Synthesis held in Autrans, France in September 1990. The set of papers selected for inclusion provides the broadest and most coherent overview of the current trends in text-to-speech synthesis. Topics include: synthesis methods, linguistic processing, prosody, system design, speech assessment and visual speech. The authors have extended and updated their contributions from the time of the conference, and specialists provide introduction or discussion of the topic addressed in each section. These papers should provide good reference and new ideas to established researchers, while encouraging people to join the ongoing adventure of basic research, development and applications in Speech Synthesis.