This study investigated the difference between peer and teacher assessment of writing compositions of advanced university students in English. Besides, it investigated the likely effect this type of evaluation might have on the learners' writing skill. To this end, 17 participants were surveyed over seven weeks. During every session, the participants wrote an essay on a topic suggested by the teacher. A blind correction was carried by their peers who graded the essays using Jacobs et al's composition profile as a descriptor grading model. Then the essays were corrected and graded by the teacher T-test was used to investigate the difference between the peer and the teacher's corrections. A questionnaire was also run to explore the attitude of the participants to such a practice, and an analysis of the participants' grades was undertaken. The results suggested that there was a significant difference between peer and teacher corrections of the compositions. The findings also revealed that the participants enjoyed such a practice as being aware of their peers' mistakes was helpful in improving their writing skill. Besides, a significant improvement in their grades was observed. The participants also reported that a combination of both peer assessment and teacher assessment had a positive effect on their writing ability.