The sound spectra of nasal murmurs in various vowel contexts have been studied by use of an analysis‐by‐synthesis scheme. The acoustic characteristics are described in terms of format‐antiformant distributions (pole‐zero locations of the transfer function for the articulatory system) in the frequency domain. It is shown that the location of the antiformant characterizes the murmur of each consonant within the class. The spectral structures are discussed with reference to the articulatory features, and acoustical interpretations of the observed characteristics of the formants and antiformant are given. Also suggested is the existence of certain gross spectral features that are characteristic of nasals as a class.