BACKGROUND Self-assessment refers to the involvement of learners in making judgements about their own learning, particularly about their achievements and the outcomes of their learning (Boud & Falchikov, 1989). Enhanced learning and professional development are the desirable end effect of self-assessment. However, questions are often raised about the reliability (i.e. accuracy and consistency) of students' self-assessment. Recent studies have revealed substantial discrepancies in students' self-assessment performance in terms of accuracy and consistency, potentially leading to the rejection of selfassessment. Hence, understanding the scope and limitations of students' self-assessment is critical. PURPOSE This study was intended to examine the reliability and importance of self-assessment, and uncover other significant learning and professional development benefits of students' self-assessment. DESIGN/METHOD Over the last three years, a longitudinal study of self-assessment of an assignment was conducted of successive first year university student cohorts, comprising of more than 75 per cent distance students. Students were asked to self-assess their assignments with the help of self-assessment guidelines and model answers. On the self-assessment feedback rubric provided, they allocated marks and provided justification for the marks for each assignment answer. Feedback from students' self-assessment was analysed quantitatively to examine the accuracy and consistency of selfassessment with respect to tutor's assessment, and qualitatively to understand the impact of selfassessment in their learning and professional development. RESULTS A comparison of students' self-assessment marks with the tutor's assessment marks showed that the majority of the students (~ 47%) overestimated their performance, while a significant proportion (~39%) remained within ±10 per cent of the tutor's assessment marks, and the remaining (~14%) undervalued their work. Correlations between students' self-assessment and tutor assessment marks ranged between-0.14 to 0.8 for various on campus and distance student cohorts. The accuracy of students' self-assessment was less than promising when considering tutor's marks as the benchmark (or expert judgement). These results are consistent with the findings of many other researchers including Boud & Falchikov (1989) and Lew et al. (2010). A wide range of correlations between various student cohorts also showed the lack of assessment consistencies. Therefore, students' selfassessment may not be a reliable (i.e. accurate and consistent) assessment technique even though some researchers have reported improved reliability under specific circumstances (e.g. Ross, 2006; Ward et al., 2002; and Rolheiser & Ross, 2006). The study has revealed that the major contributor to learning during students' self-assessment is not the accuracy or consistency of students' selfassessment. Instead it is the change in students' psychological perspectives when they go through the process of self-assessment. In this study metacognition, constructivist learning, and self-efficacy have been identified as triggers to students' learning and professional development during self-assessment. CONCLUSIONS The study has demonstrated that students vary considerably in assessing their own work. Hence, the learning gained during students' self-assessment may not come from the accuracy and consistency of self-assessment. Instead, it is triggered by the change in students' psychological perspectives resulting in metacognitive knowledge, learning constructivism and development of self-efficacy.