The primary uses of computers in the construction industry have been shifting, over the past four decades, from the evaluation of proposed design solutions, to their graphical (and other) representation, and more recently to facilitating collaboration among the various professionals who are involved in the design process. The paper argues that what may appear to be shifts in emphasis actually represents convergence on a single, original goal: the use of computers to help designers assess the quality, desirability, and the implications of their creations. The paper shows how the formerly independent components can be joined into an integrated collaborative design environment, where they build upon and strengthen each other. Moreover, the paper argues that this convergence represents the future of CAAD research and development, providing the appropriate answer to the upcoming needs of the construction industry, whose products have become too complex and must abide by too many requirements for any one professional to handle all by himself. The paper argues that further improvements in the overall quality of the products, and the process of their design, will only accrue when the heretofore separate solutions are considered together, as integral parts of an overall solution. The paper describes the efforts that have been made by the CAD Research Group in Berkeley over the past six years in developing an integrated collaborative design environment that can facilitate multidisciplinary, a- synchronous design of buildings. The environment includes several semantically-rich, shared product representations, a network of distributed evaluators, and graphically enhanced collaboration and negotiation tools.