Genre, defined as a socially accepted way of using language within a particular discourse community (Hyland, 2007), has had a significant influence on the way in which learning to write is conceptualized at university. Learning to write was conceptualized as a means for developing genre knowledge. Therefore, this study is intended to find out whether students develop genre knowledge implicitly or explicitly as they move from one level of study to the next. To meet this objective, the data consist of 34 introduction sections of research articles by students in TESOL master’s program at Ecole Normale Superieure du Burundi. Seventeen introductions were written by first year students (batch 2024-2025) who were taught academic writing through genre exemplars and 17 introductions were written by second year students (batch 2023-2024) who were taught academic writing through process approach. To analyze the data, these introductions were rated by experienced lecturers in the department of languages and social sciences in term quality and genre feature. The results showed that there was a difference in genre knowledge and a significant difference in writing quality between Master I and Master II students. It was revealed that master I students, who were taught academic writing through an explicit use of text exemplars, were rated higher on genre knowledge and writing quality than were master II students, who were taught academic writing through an implicit use of text exemplars. These results have a far-reaching implication for teaching academic writing to university students in Burundi.