Currently, there is a great debate concerning whether peer review is an effective activity in the university-level English-as-a-Second-Language writing classroom. Peer review offers the unique opportunity for second-language writers to share their writing, evaluate others' work, and discuss their observations and opinions about writing in an authentic environment. This study investigated the interactional dynamics of peer review sessions in university-level ESL writing classrooms. The spoken data (transcripts of peer review sessions) were evaluated to determine the types of interactions that occurred and their functions. The written data (rough and final drafts from each participant) were evaluated to determine the quantity and quality of the changes that the participants made. The two data sets were then compared to determine whether (and to what extent) the peer review interactions led to improvements on the drafts. The data indicated that suggestions made during peer review correlated to positive changes if the participants negotiated the suggestion, and if the suggestion pertained to global-level issues in the paper. While the data showed that participants preferred to make changes unilaterally, it also indicated that peer-reviewed suggestions correlated with a higher percentage of positive changes than writer-initiated suggestions. Further, the data indicated that peer review was particularly favored by those participants who had no previous experience with this activity. These results indicated that peer review is an effective activity, especially for students who are new to it. It is best framed as supplementary to the student's existing writing process. Future research should focus on triangulating the data with post-activity student interviews, in order to corroborate the results.