Speech recognition systems have been applicable in wide areas as various speech recognition methodologies, techniques and tools have been developed and implemented to generate a natural and intelligible speech. In this regard, this work attempts the possibility of developing a prototype speech recognition system for Sidama language using Hidden Marcov Model. The study has conducted extensive study on the language features, the components, speech recognition tools; the techniques used in speech recognition design, and identified those component that are dependent on the characteristics of language. Finally this work has showed a working prototype speech recognizer for the language, tested the performance of the system and compared its accuracy, and recommended measures for similar researches and projects. This work, therefore, will be useful to researchers, Speech application developers, Educators and other individuals or institutions working on similar projects.