Numerous authors have presented ontology building tools that have all been developed as part of academic projects and that are usually adaptations of more generic tools for geo-spatial applications. While we trust that these tools do their job for the special purpose they have been built, the GIScience user community is still a long way away from off-the-shelf ontology builders that can be used by GIS project managers. In this article, we present a comparative study of ontology building tools described in some twenty peer-reviewed GIScience journal articles. We analyze them from the perspective of two application domains, crime analysis and transportation/land use. For the latter, we developed a database schema, which is substantially different from the three main templates commonly used. The crime analysis application uses a rule base for an agent-based model that had no precursor. In both cases, the currently available set of tools cannot replace manual coding of ontologies for use with ESRI-based application software. Based on these experiences, we outline a requirements list of what the tools described in the first part of the article are missing to make them practical from an applications perspective. The result is an R&D agenda for this important aspect of GIScience.