At Halmstad University we got off to a very slow start when it came to self-archiving of our scientific publications. Because of the library’s involvement in the management of student papers, it was a fairly simple process to move from the storage of bibliographic data to fulltext archiving of these papers; however, it was a completely different matter with research publications. There was nothing to build on as no database or collection of any sort existed of the research output. Thus we had to start from scratch, which meant convincing researchers, the heads of research departments and others of the potential benefits of setting up an open archive. Despite intensive lobbying from the library, it was only after a decision made by the Head of the University, stipulating that all researchers were responsible for registering their publications in the system, that we experienced an increased activity in terms of registrations and questions concerning self-archiving. This confirmed, what we had already heard from other experiences, that although it would be good if the researchers could see the benefits of voluntarily registering their publications, some sort of mandate was still necessary to get things in motion. As however the decision did not mandate fulltext archiving, this has led to a system containing very few fulltext documents. The coming challenge lies in increasing the amount of fulltext in the system, and in particular, of refereed scientific articles.