Over the past few decades, corpus-based investigations have contributed greatly to our understanding of academic discourse. One important domain of academic language use, however, has yet to be fully explored from a corpus-based perspective: textual borrowing. Though it is widely recognized that much of what we write in the academy is in some way based upon what has been written before, little is known about when, how often, and in what ways academic writers re-use the language of others. In this paper, I describe my own attempts to provide corpus-based descriptions of student paraphrasing, I highlight the ways in which this research has challenged assumptions about student source text use, and I outline possible directions for future textual borrowing research