This study was designed to determine the value of using a mirror for speech teaching. Data derived from manner, place, voicing, and blend errors produced by 27 hearing-impaired children while imitating consonant-vowel syllables under three conditions (audition alone, audition plus direct vision, audition plus vision using a mirror) were analyzed to determine whether the use of a mirror enhanced or detracted from the accuracy of their responses relative to the other two conditions; whether different types of errors were made under any of the conditions; and whether error types under each condition were related to age or hearing level. Although more place errors occurred under the auditory condition, the difference between the numbers of errors made under the two auditory-visual conditions was not significant. The use of a mirror neither enhanced nor detracted from the accuracy of responses of subjects in this study. This finding cast doubt upon the value of using a mirror as a general aid in speech teaching.