While deaf individuals often develop speech that is usable for communication, their speech is often judged different from that of their normal-hearing peers. The purpose of this study was to examine the acoustic characteristics of intelligible speech by deaf children and to compare these characteristics to those of hearing speakers. A corpus of 32 words [after Kent et al., JSHD 482--499 (1989)] was obtained from ten deaf and five hearing children. Utterances were subsequently judged by a panel of listeners in order to select those productions by both the deaf and hearing speakers that were understandable. The intelligible utterances were then examined acoustically. Among the measures explored were formant patterns, frequency attributes of fricative consonants, temporal attributes of segments, syllables and words, and voice onset time. The resultant measures were compared across the two groups of utterances (intelligible deaf and intelligible hearing). The results showed some similarities and some differences between the two groups of utterances. The implications of these findings on speech intervention, and on the use of automatic speech recognition with disordered speakers will be discussed.