In most industries between 70% and 90% of all factors that can potentially cause an accident refer to human error. Human error research in industrial high‐risk socio‐technical systems has a long tradition, but is relatively new to medical settings where it started to become more important just two decades ago. Despite lacking official statistics, 70%–80% of all incidents and accidents in the medical sector have been attributed to have been triggered by human error. In light of these alarming figures, the present research aims to identify sources and causes for human error in the medical sector. The conceptual framework of the present study is based on Gordon Dupont's Dirty Dozen Model which includes twelve error categories for human error in aviation, covering organisational and social aspects but also considers individual factors which can provoke human errors. For the current study, 18 qualitative interviews based on the Critical Incident Technique were conducted with medical doctors, medical‐technical assistants and nursing staff. The interviewees had to retrospectively describe specific events which led to critical incidents or situations. Based on the Dirty Dozen Model the interviews were analysed with regard to the mentioned causes that provoked errors. After the sources of error have been identified and extracted from the interviews, the second step was the classification of possible combinations of different error sources, since human error often results from the interplay chaining of multiple error sources.