At the undergraduate level, considerable evidence exists to support the use of peer assessment, but there is less research at the graduate level. In the present study, we investigated student perception of the peer assessment experience and the ability of graduate students to provide feedback that is comparable to the instructor and that is consistent between assessors on a written assignment. We observed that students were very supportive of the activity and that negative concerns related to inconsistent peer evaluations were not supported by the quantitative findings. Our quantitative analyses showed that the average grade of the student reviews was not significantly different from the average grade given by the instructor, although student reviewer reliability was not high. Students showed a significant grade improvement following revision subsequent to peer assessment, with lower graded papers showing the greatest improvement; greater grade change was also associated with an increased number of comments for which a clear revision activity could be taken. Design of the graduate peer assessment activity included several characteristics that have been previously shown to support positive findings, such as training, use of a clear assignment rubric, promotion of a trusting environment, use of peer and instructor grading, provision of directive and non-directive feedback, recruitment of positive comments, and use of more than one peer assessor. This study, therefore, builds on previous work and suggests that use of a carefully designed peer assessment activity, which includes clear direction regarding actionable comments, may provide students with useful feedback that improves their performance on a writing assignment.