The study explored the pedagogical use of Text-To-Speech (TTS) synthesizers by comparing two groups of students learning the pronunciations associated with Regular Past Tense (RPT) marking in English (e.g. talk[t], clean[d], want[id]). While one group received TTS-based instruction, the other was taught RPT with teacher assistance. Our findings indicate that both groups behaved similarly regarding the production of RPT, and that the target allomorphs are acquired in the following order: id>d>t. [For the complete volume of short papers, see ED590612.]