The master thesis represents an essential component of a postgraduate programme. First of all, the student has to define a topic and a set of relevant research questions. After having settled the supervision agreement, he or she can embark on the actual work on the thesis. Nowadays, a wealth of preceding publication output is accessible and it is tempting to "re-use" this material. If time is short and deadlines beckon, plagiarism may seem to be the easy way out. This means that (substantial) chunks of already published work is reused in the master thesis, but even paraphrased modifications may still be in dangerous proximity to the original publication. Moreover, an author may consider to incorporate large parts of his/her own previous work, thereby engaging in self-plagiarism. The supervisor has a crucial role in this regard. If he or she has an excessive number of students to supervise, there may not be enough time for individual consultations. On the other hand, supervisors may expect to see work in progress on which to give feedback, which should then in turn be reflected in the final version. If a more or less final version is submitted without intermediate stages, this might be regarded as a warning sign. Another tell-tale signal is a mismatch between a student’s oral proficiency of expression versus "perfect" writing. There are several tools to check digital documents for any degree of plagiarism. These reports need to be critically checked, but they can provide a helpful first impression. It also has to be noted, that retrospective digitisation may lead to the detection of plagiarism in the future. The contribution will conclude with some comments on ghost writing.