This study investigated the effect of transcribing group-discussion tasks on the development of university students' autonomy and oral proficiency. A quasi-experimental research design was followed to compare the performances of four groups: two experimental groups and two control groups (each group divided into low and high proficiency students). The study lasted for 12 weeks, and the teacher assigned a classroom oral discussion task in each session. The students were divided into discussion groups of three or four students, with low and high proficiency learners in different groups. The participants had to record their group discussion tasks. Control groups’ students had to submit their recorded conversations to their instructor, but they did not do any post-task activity. However, the experimental groups’ students had to transcribe their recorded speaking tasks, to find their own and their peers' grammatical mistakes, and to correct them. Finally, while working in groups, students discussed the texts and reformulated their mistakes. Employing ANCOVA to analyze the results, researchers found that experimental groups significantly outperformed the students of the control groups on post-tests of oral proficiency and learner autonomy. Thus, transcription followed by reflection on inaccurate production contributed to the superior performance of participants in the experimental groups.