In this paper a Large-Vocabulary, Real-Time, Isolated Word Recognition system is presented. While the final goal of the project is that of recognizing words from a vocabulary whose size is of the order of 10-20 thousands, the present system is intended to perform real-time recognition on vocabulary subsets (cohorts) of up to 2 thousand words. The system is implemented on an EMMA2 multiprocessor for a fast response. Basic features of the system are the use of sub-word units (diphones) for the acoustic measurements and the derivation of synthetic symbolic word templates directly from the required lexicon.