ABSTRACT During the academic year 1973–74, the authors initiated an experimental program in first-year French at Dalhousie University. A self-instructional, programmed text was used, and performance objectives were set for each unit of work. Students were responsible for their own learning, proceeded at their own rate, and were allowed repeated attempts to meet the minimal requirement of 75% mastery. The role of the instructors was to offer individual help where needed, usually in the form of additional explanations and controlled practice, to evaluate each student's progress through frequent quizzes for listening comprehension and separate tests for oral and structural accuracy, and to be available for informal conversation in French. Regular ‘communicative activities’ were required of all students in which they were expected to express themselves entirely in French on a variety of topics and in a variety of ways, some suggested by the instructors and others designed by the students themselves. Comparisons with control groups at the end of the year indicated a significantly higher level of performance for the experimental students but considerably less material covered. Student attitudes toward the experimental conditions, as evidenced in the results of a detailed questionnaire, were judged to be encouragingly positive. No general conclusions can be drawn without oversimplification, but the tentative results indicate that, under appropriate conditions and with suitable students, this type of learning can be an effective alternative in foreign language education.