This paper compares students' judgements on scores for their written assignments with the marks awarded by their tutors on professional programmes at a higher education institution in the United Kingdom.Assessment is a key and influential element in students' learning experiences. Indeed, an effective way of altering the form and type of teaching and learning is to change assessment practices. This is because, largely through the encouragement of the Quality Assurance Agency, teaching and learning and assessment have become more 'constructively aligned'.There is detailed consideration of self-assessment which is becoming increasingly common in higher education institutions. The reasons for its greater use are related to both educational considerations and the need for staff time to be used as effectively as possible. It is claimed that educational reasons for the increased interest in self-assessment include its ability to empower students by increasing their confidence through a far clearer understanding of the assessment criteria and the assessment process. Self-assessment also improves their understanding of their work as it greatly increases reflection on what they have written. It is argued that students take more responsibility for their learning and this increases satisfaction levels, makes them more independent and consequently less dependent on the lecturer. Reference is made to work on self-assessment by writers such as Boud, Falchikov, Brew, Longhurst, Stafani, Dochy, Segers and Sluijsmans. These writers discuss issues such as definitions of self-assessment, its strengths in terms of how it can enhance learning and the problems associated with its implementation. This literature is analysed critically and synthesised to provide the theoretical underpinning for a small-scale study.The study involved taking random samples of recently appointed higher education lecturers on a Postgraduate Certificate in Learning and Teaching programme and students on pre-registration and post-registration nursing programmes. Students were asked in individual semi-structured interviews to state the score that they thought they would achieve on an assessed essay and justify their decisions using the criteria on the marking grids for their particular programmes. The score predicted by the student was then compared with the actual score awarded by the tutor. Simple statistical analysis was conducted by comparing the mean of the students' predicted score with the mean of the actual score awarded by the tutors in a range of marking categories; 40-49%, 50-59% 60- 69% and 70% and over. The difference between the two means was then illustrated graphically for the three groups of students.Through reference to relevant literature, explanations were sought to explain why these different groups varied in their accuracy in predicting their scores. It is concluded that a clear understanding of the assessment grid lexicon helps students to compare their work accurately against the assessment criteria. Also, knowledgeable students in terms of both academic experience and employment are more confident and therefore less likely to underestimate their abilities. Their judgements of the scores are likely to be similar to those of their tutors. On the other hand, 'novice' students with limited academic and work experience are inclined to score themselves lower than their tutors