Assessment loads have increased due to significant increases in student numbers and a relative stagnation in staff numbers. Assessment of students′ work is now a significant issue in many universities in the UK. This potential problem is compounded by increased consumer pressure, either directly from students or else from parents. One response to this challenge has been the increased use computer or machine-marked examinations. In many universities, however, final year students tend to be assessed using essays and other forms of structured writing, including library projects and dissertations. These are time consuming to mark and students frequently claim to be either ignorant or unsure of the marking criteria. In Bristol we have examined peer-assessment of final year essays, in the Physiological Sciences degree program, as a vehicle to manage the assessment burden and also to provide students with insight into how their written work is assessed, including how grades are determined. Results indicate that students are able to grade essays to within ±5% of the mark awarded by experienced teaching staff. In each of 2 years, an essay title was made available to students (n= 25 and 22) 7 days prior to the assessment. The essays were then written under examination conditions and scripts identified only by candidate numbers. After a break of 30 min, the marking scheme was developed collaboratively by the students in a process that was moderated by two members of staff. Each student was then allocated an essay from the group at random and was required to grade the essay according to the agreed marking scheme and some additional guidelines regarding structure. The mark had to be justified by the inclusion of three comments about the positive aspects of the essay and a further three comments focussing on those aspects that could be improved upon. Each marker identified themselves on the script using their candidate number. Because each student received a mark for their work and allocated a mark to the work of another, it was possible to look for the existence of any correlation between the mark received and the mark given. In each year a correlation was found to exist between the mark a student received for their work and the mark they awarded to the work of another student. Using graphpad Prism (version 4.0), a correlation analysis indicated the presence of a correlation (P<0.0001, n=22) and that approximately 45% of the variation in the mark given could be accounted for by the mark received. Anonymous questionnaire comments from students were positive. Many found the opportunity to see how another student had approached the question and interpreted the same areas of the literature invaluable. Several students openly stated that it caused them to reflect critically on their own essay, something they had not tended to do previously. In summary, peer-assessment of final year written work benefits both students and staff.