Big data and data-driven innovation are drivers for economic growth. To capture this growth, data often need to be shared among organisations. However, many challenges to sharing data among organisations exist. This paper investigates how governance is organised in inter-organisational data collaborations. First, based on literature, four archetypical modes of governance are identified: Market, Hierarchy, Bazaar and Network. Subsequently, these theoretical modes are investigated empirically by exploring governance modes in four use cases. Based on a cross-case comparison, we find that major challenges to data sharing are the commercially sensitive nature of data and privacy risks. Due to legal implications, sharing of personal data always takes place hierarchically. Therefore, coordination and control over data need to be firmly in place before organisations engage in data sharing. Further research should look into how these aspects can be organised in inter-organisational data collaborations to foster innovation.