Protected cultivation allows the production of crops under very diverse conditions. The afferent technologies, however, can be complex and costly. Hence, the effectiveness of greenhouse systems, both from economical and technical points of view, depends more and more upon the availability of sophisticated, on-line and automated decision-making systems capable of dynamically optimizing the underlying processes. Such systems have to be able to intervene in a number of areas such as crop protection, climate control, crop nutrition, and operational and strategic planning. In other words, the automated control system that takes charge, in part or in whole, of a greenhouse production system needs to draw upon a considerable body of knowledge in order to be effective. In this article, a conceptual and operational framework (GX) supporting the representation and management of real and virtual entities used or found in greenhouses is described. It was designed to support various types of digital process controllers as well as the creation and deployment of knowledge-based control strategies. GX makes use of the object-oriented paradigm as expressed in the Smalltalk programming system. It has been used as a simulation platform and for the real time monitoring and control of a greenhouse range. The GX programming environment is extensible. It can accommodate a wide variety of situations and provides a semantically rich environment for the design and operation of knowledge-based greenhouse control systems.