In this paper based on the needs of speech application research such as speech synthesis and speech recognition, combined the characteristics of the Uyghur language, for the first time, established the "Uyghur speech acoustic parameters database" which includes common sentences, words, syllables and phonemes and their acoustic parameters. And the details of key technologies in the procedures of establishing database such as text selecting principle, speech labeling rules and labeling, database structure, acoustic feature selection, parameters acquisition are mentioned in this paper as well. The results of this paper have high value on speech research and application development of Uyghur even the other languages of Altaic.