In spoken syllables such as "ta," the interval between the release of the tongue constriction for the stop consonant /t/ and the onset of the vowel is called voice onset time, or VOT. Voice onset time is an important determinant of whether the initial consonant will be heard as a /t/(values of 60-90 ms) or as a /d/ (values of 0-30 ms). VOT information, immediately following a spoken syllable, can provide a speaker with feedback for modifying speech production. Such information can help the hearing-impaired learn to speak. It may also help people who learn English as a second language, since they often produce /b,d,g/ and /p,t,k/ with inappropriate VOT values. A prototype portable device measures VOT for initial voiced and voiceless stop consonants (e.g., "da" and "ta"). A dual-microphone method is used for acoustic measurement. A microphone in front of the mouth picks up the radiated acoustic signal; another over the larynx transduces vocal vibrations that mark the beginning of a vowel. Analog circuits process the transducer signals and provide gain and filtering. Filters were designed on the basis of the acoustic properties of stop consonants. The output from each analog circuit is fed to a comparator that compares the signal level with a fixed threshold voltage reference. A digital timer starts when the amplitude of the oral signal voltage exceeds a threshold and stops when the laryngeal signal voltage exceeds a threshold. VOT values obtained by the device were compared with those made from digital waveforms of words spoken by five talkers.(ABSTRACT TRUNCATED AT 250 WORDS)