Abstract. In recent years, controlled vocabularies have become available on the Web using SKOS 1 , i.e. they are linked to each other in order to be used in an interoperable way. Well-crafted controlled vocabularies are beneficial for, e.g., search and retrieval systems that provide functionalities like search term completion, query expansion or the ability for inter-domain queries. Some of these vocabularies are created collaboratively by experts, holding expertise in different domains. In order to support vocabulary contributors to create high quality vocabularies, we propose a method that semi-automatically ensures vocabulary quality in collaborative authoring processes. The proposed approach tackles this issue by (i) defining a set of criteria that serve as a metrics to measure vocabulary quality and (ii) introducing a method to continually assess and improve this quality. As a result of our approach, the developed vocabularies are expected to better fit the intentions of the contributors and are more useful for reuse and adoption on the Web of Data.