Speech-evoked, auditory event-related potentials or simply the speech evoked potentials described by many contemporary researchers can provide informative value regarding the neural mechanisms that underlie speech processing by the human auditory system. Nevertheless, there remains a lack of physiologic and objective data in human speech perceptual study, especially within the Korean language which differs in phonological characteristics and in speech sound acquisition and developmental patterns for the phonological processes. As future of evoked response audiometry, we need to revisit classic studies of the speech perception (i.e., Miller & Nicely, 1955; Bilger & Wang, 1976), apply to formula of speech perceptual similarity and perceptual dimension, and extend to then measure objective test using oddball paradigm for supporting behavioral perceptual data, which can be a way of building bridge between behavioral and physiological findings. In this lecture, we will focus on how Korean listeners perceive speech sounds at the level of consonant in the behavioral and physiologic ways using confusion matrix and mismatch negativity, respectively. The results would be helpful for therapeutic approach to hearing-impaired listeners in terms of confusable sounds classification in the audiologic rehabilitation.