Over the past decade, peer assessment has become a controversial concept in higher education. There is much research evidence highlighting the effectiveness of peer assessment and its' potential advantages including the promotion of learners' responsibility, metacognitive development, autonomy, and deep learning. On the other hand, significant research findings point to its ineffectiveness in improving students' marks and performance, poor quality feedback, and potential bias. This paper presents a conceptual overview of the literature on peer assessment with particular emphasis on prior empirical findings. A case study with practical evidence from the author's context is then presented, reflecting on the implementation of peer assessment in a postgraduate unit over a two-year period. This sets a preliminary means to investigate whether the problem is with product (ie. peer assessment is fundamentally and conceptually flawed) or process (ie. peer assessment is wrongly implemented). The paper concludes with recommendations for future research and practice.