Until today, the notion of trust in multilingual conversations has been treated merely in passing in translation studies. Only recently a few authors focused on this concept and showed its relevance for research and practice. This thesis aims to investigate how trust is built in interpreted communications, especially in the field of community interpreting. Do people put trust in interpreters? Are the interpreters able to influence this process? Do they have a kind of power position within the conversation and/or how does the presence of power influence trust? In which ways are basic aspects of the professional work life such as neutrality, loyalty and professionalism linked with trust? What is the difference between the building and displaying of trust towards professional and amateur interpreters? Hypotheses on building and maintaining trust, mistrust, hierarchy and power in interpreted conversations, as well as the often diverse role of interpreters are explored through semi-structured interviews with community interpreters. The analysis of the data revealed that interpreters are highly trusted and ? apart from exceptional cases ? are not at all considered a necessary evil in the interaction. Furthermore, concepts like neutrality, loyalty and power mostly depend on the participants in the interaction as well as on the different settings, and are rated as essential for the building of trust. In addition, the results show that professional interpreters are more trusted than amateurs because of their competences and their professional attitude, especially within the medical context or in delicate situations. All participants of triadic conversations usually consider trust as relevant.