The fricative sounds are produced when the vocal tract is excited by acoustic noise generated from turbulent air flow near a narrow constriction in the tract. The mechanism of the production of fricative consonants is investigated with the aid of an idealized mechanical model of the human vocal organs. Power-density spectra as functions of air flow are calculated from sound pressure measurements around the model. An electrical analog circuit is derived to represent the model for sound radiated on axis. “Source” spectra are inferred from comparisons of the transfer function of this electrical circuit with the measured data. On the basis of the model studies, the constricted vocal tract is approximated by a simple, lumped element circuit excited by appropriate noise sources. A comparison is made between the spectra of the outputs of this system and existing data on the natural fricatives /ζ/, /s/, and /f/. Listening tests performed on the synthetic sounds in isolation and in syllables constructed with synthetic vowels having appropriate formant transitions have demonstrated the adequacy of the analog circuit. In agreement with other studies, distinction between /θ/ and /f/ is made principally on the basis of formant transitions rather than spectrum. [This work was supported in part by the U. S. Air Force (Air Force Cambridge Research Center) and the U. S. Navy (Office of Naval Research).]