One of the most serious limitations of the Entity Relationship (ER) Model in practice is its inability to cope with complexity. Once data models exceed a certain threshold size, they become difficult to understand (end user’s viewpoint) and also to document and maintain (analyst’s viewpoint). A number of approaches have been proposed in the literature to address this problem, but so far there has been no systematic empirical research into the effectiveness of these methods. This paper describes an action research study in which a method for representing large ER models was tested in a large application development project in one of Australia’s largest commercial organisations. The research was successful in achieving both practical and research outcomes􀁿it resulted in change of data modelling practices in the organisation, and the method was refined significantly as a result of experiences in practice. However a major problem experienced in this study was that the size of the project imposed constraints on the evolution of the method. Because of the number of people involved, it was difficult to make changes to the method “on the fly” and to experiment with variations of the method, as is customary in action research.