The phonetic transcription ability of 78 college students whose transcription instruction was administered by means of pre-programed Language Master cards was compared with that of 81 students whose instruction was non-automated. Ability was measured by seven weekly tests. There was no significant relationship on any of 29 variables with type of instruction. Intercorrelational techniques showed no positive correlation for sex, but positive correlations of grade point average and transcription and theory tests, and in four of the six Seashore Measures of Musical Abilities subtests (timbre, memory, pitch, and time). On questionnaires, students with live instruction indicated that they were significantly more satisfied (p=.05) and the main reason given was the feedback obtained from verbal imitation and the instructor's immediate critical reaction. It was concluded that live instruction be supplemented by machine practice. Three references are cited; word lists, instructions for Language Master users, satisfaction scale, grade data, and questionnaire data are provided. (Author/SN)