State-owned enterprises (SOEs) provide important services such as power supply, postal services and public transportation. In doing this, SOEs operate at the intersection of political/regulatory and economic expectations. Therefore, SOEs are commonly considered in the academic literature as hybrid organisations. Hybrid organisations in the public sector seem a way to combine the best of two worlds: high-quality, broadly accessible public services and affordable prices thanks to commercial efficiency and flexibility. However, this hybridity has an important implication: the accountability of SOEs is increasingly complex and blurred. To whom, for what and by which standards are SOEs accountable? Previous studies on these questions have been mainly interested in the differences between the accountability of SOEs to private and to state actors. However, this focus neglects that the state is not a monolithic actor, but the state assumes several roles with regard to SOEs. In each of these roles, the state has expectations what SOEs shall and shall not do, and defines respective accountability criteria. This may lead to incoherences between different expectations of the state and the respective accountability criteria. Hence, research question 1 is as follows: How do the multiple roles of the state vis-à-vis SOEs lead to incoherent expectations and accountability duties? Based on this analysis, research question 2 is as follows: How do SOEs manage incoherent public expectations and accountability duties vis-à-vis the state? In order to analyze the research questions, six case studies of Swiss SOEs from different network industries are conducted. The cross-case analysis resulted in typologes for each of the two research questions. The typology for research question 1 shows the importance of differentiating between expectations concerning goals and means, positive and negative expectations and social/political and financial expectations. The typology helps to understand how incoherences arise when the different types of expectations of the state are not aligned. The typology for research question 2 provides a systematic overview on how SOEs deal with incoherent expectations and accountability duties. Three groups are differentiated, with each of them comprising several practices: prevention, partnerships and participation. Practices are grouped along the two dimensions ¿issue-specific/short-terms versus issue-spanning/long-term¿ and ¿compliance-orientied versus impact-orientied¿. The results add to existing literature by providing analytical typologies which can be applied and refined in further studies on SOEs in network industries. More specifically, the typology for research question 1 contributes to the political science literature on the governance and regulation of SOEs, whereas typology 2 enrichs management science literature on the management of hybrid organisations. Practitioners in both public administration and SOEs may benefit by using the typologies in order to identifiy and discuss incoherences and possibilities of coping with incoherences in a systematic way.