It has been shown by Flanagan and Landgraf [IEEE Trans. Audio Electroacoust. 16, No. 1, 57–64 (1968)] that in the transformation between glottal area and glottal volume velocity during voiced vowels, the effect of the interaction between the acoustic impedance of the glottis and that of the supraglottal vocal tract cannot always be neglected. When there is such interaction, the glottal-supraglottal acoustic system must be considered a nonlinear system with time-varying parameters. Thus, any inverse-filtering process designed to extract the glottal area waveform from measurements of supraglottal pressure or air flow must also be nonlinear. This paper reports initial efforts to determine the conditions under which a nonlinear inverse-filtering process exists that will yield a function directly related to glottal area from recordings of oral volume velocity, and how such a nonlinear inverse filter could be implemented.