SRS (Speech Research System) is a highly flexible computer system for the development of text-to-speech rules for any language. Its powerful interactive facilities and special rule language were designed to facilitate rule development by programmers and nonprogrammers alike. The system has been used for instruction in acoustic phonetics, for the generation of stimuli for [human] perceptual experiments, and for the development of synthesis rules for a variety of languages. The SRS rule framework, which allows users to express 4 kinds of rules that apply in succession to convert text to sound was studied. The user-oriented nature of the SRS interactive facilities are illustrated, and the system''s implementation and use are described.