The purpose of this study was to specify variables which function significantly in the racial identification and speech quality rating of Negro and white speakers by Negro and white listeners. Ninety-one adults served as subjects for the speech task; 86 of these subjects, 43 Negro and 43 white, provided the listener responses. Subjects were chosen to provide a sample approximately representative of the distribution of socioeconomic status scores in the southeastern United States. Listeners were asked to identify the race of each speaker and make a speech quality rating of recorded samples . The Articulatory Product score developed by Guttman was used as an independent, semi-objective index of speech proficiency. In addition, a spectrographic analysis was carried out using a sample consisting of ten Negro male and ten white male subjects. All speakers used in this analysis had been correctly identified by listeners as to race 9 5 percent of the time or better. Results were as follows: (1) The number of phonetic distortions by speakers predicts racial identification. (2) Socioeconomic status score and Articulatory Product score predict speech quality rating of speakers by listeners. (3) No significant intergroup differences were found on spectrographic variables. Negro speakers used in acoustic analysis, however, had consistently greater attenuation of format amplitudes [u] vowel than white speakers. (DO)