A procedure for evaluating the operation of two electronic formant extractors is described and results of the evaluation are given. The formant extractors are designed to yield electrical outputs representing the formant frequencies of continuous input speech. These outputs are compared in a prescribed manner to formant data derived from sound spectrograms. The results show that the more reliable extractor of the two follows the first formant of vowel sounds in speech within ±150 cps greater than 93% of the time; it follows the second formant within ±200 cps greater than 91% of the time. The first and second formant outputs of the same extractor are simultaneously within the above tolerances greater than 85% of the time.