Abstract A trivial information system consists of three components: a user, a container for the information and a tool to manage the information. The focus of this research is more complicated with several users and organisations, with several databases in those organisations and with an inter-organisational information system. In inter-organisational information systems information flows across boundaries and the boundaries as well as information itself must be carefully taken into account. This research takes seriously the users who represent different professions and who were more or less involved in the information system implementations. In a distributed project, also the container is under special attention. That is discussed in the dissertation, as well. The approach in the research is mainly interpretative and subjective. The empirical material comes from several cases but three of them acted as a background for the main case that was conducted in an academic environment. The case appeared to be versatile due to its many stakeholders and it offered several topics to be explored. The goal of the research was introduced in a meeting to the attendees: "The aim of the thesis is to describe how a joint information system of different universities was finally implemented or why it wasn't." This introduction predicted challenges that were to be overcome during the subsequent years. In addition to the versatile case, also the information system implementation offered fruitful viewpoints for the research. Inter-organisational information system projects are described in literature to some extent. This case contributes the research with the many organisations and users that were involved in the development project. Furthermore, the information system was planned to be built on ready-made specifications and it increased the challenges to succeed. This led to the lesson that a new model to be used in information system developments is needed because existing models do not support this kind of development process. The research supports the concept that action research and design science complement each other especially in development projects when the output is an information system that has been implemented in collaboration with several users and stakeholders. The results of this thesis emphasise the need to manage potential competition between participating organisations, lack of interaction between parties and security and user authentication. Proper project management is needed with active and confidential terms between all project members. In addition, flexible working routines are found workable.