This thesis provides an extensive overview of the issues involved in adopting a knowledge based approach to the development of an integrated process design environment. The majority of the work described requires a good understanding of topics from several research areas outwith mainstream process engineering. The topics covered include logic, graph theory, knowledge representation, and theories of comprehension in cognitive science. Introductory reviews of these subjects are provided in order to make the thesis understandable to an engineering audience. The thesis consists of four main sections. The first section consists of a critical review of previous attempts to develop integrated design systems using database technology. The section highlights the problem of the use of such systems in terms of data and knowledge integrity. A novel AIP (Appropriate and Incremental Parallelism) systems engineering framework is proposed along with a long term research framework to investigate the use of knowledge representation techniques to implement this design methodology. The second section provides a review of the basic aspects of graph theory, logic and production systems that are essential to understanding the rest of the work. The third section describes the development and use of rule based techniques, blackboard systems, and object oriented/frame based systems. A novel representation system, (Combined Logic and Procedures - CLAP), is described that is specifically designed to represent the various types of information needed in a process design environment. Drawing on these developments, the fourth section provides an extensive review of the many problems associated with the use of production and frame based systems. This is followed by a review of more advanced ideas on theories of comprehension in cognitive science. In particular the study of conceptual categorisation and mental models is discussed in some detail. The preliminary development of a system, Designer's Assistant, is then described that attempts to embody these ideas in a working system. The main conclusion of the work is that current knowledge representation techniques are far from adequate in terms of being able to provide the tools needed to develop an intelligent design environment. The work described in the latter half of the thesis indicates that considerable research remains to be done.