This paper discusses a three-year experiment to build a distributed research group, equipped with state-of-the-art computing facilities, spread over three cities in Australia. Despite the provision of the sorts of facilities to be expected in cooperative buildings, such as high-speed networks and videoconferencing, significant synergy (i.e., closely-coupled collaborations) among the distributed subgroups did not develop. This was not only due to the problems of distance, but was exacerbated by several political and organizational issues. An important lesson is that successful ‘cooperative buildings’ will depend not just on the technology but also on an appropriate managerial, organizational and political climate in which these resources can be meaningfully exploited. The paper outlines the experiment, discusses why synergies did not emerge, and points to implications for cooperative buildings and design paradigms based on the notion of pattern languages.