The problems associated w i t h the process of educational software design are complex and need to be considered from a number of different perspectives. In this study a number of factors are identified as contributing to difficulties generally experienced by software designers. It is suggested, however, that the factor which underlies all others is ineffective or inefficient communication. As the design of Educational Software Systems is a complex, multidisciplined process, the communication of primary interest is that between between experts from different disciplines. To help focus on such problems and processes most discussion is in terms of two representative experts: a Teacher or Educator, and a Computer Programmer or Systems Analyst. In the first chapter the complexity of the task of categorising and evaluating information about educational software is discussed. A need is recognised for some form of conceptual construct which would allow direction and progress in software design to be determined. The concept of a continuum between the &quoteComputer as Tool&quote and the &quoteComputer as Tutor&quote is introduced as a logical basis for such a construct. In this and several other chapters the focus is on the design of Intelligent Educational Software, while not intending to imply that this is its only useful or desirable form. If, however, the design of Intelligent Educational Software is better understood, the designing of less complex forms of software should become much easier, and will make possible the teaching of Educational Software Design as a topic for formal study. The second chapter addresses the problem of interpersonal communication between experts in different disciplines who have no common technical language. The design of educational software is made more difficult by the fact that teachers find it difficult to describe &quotewhat they do&quote when they teach. The concept of a language of accommodation is introduced and discussed. The general problem of software acquisition, design management, and evaluation is addressed in chapter Three. The interaction between the roles of Educator and System Analyst is considered in relation to the types of software available today. It is suggested that collaborative design between experts from different fields can be described and analysed as a set of complex learning behaviours. The process of design is recognised as a learning process which, if better understood, can be improved and taught. Chapter four considers the problem of human/machine interaction. An operational model, or designer's check list, to aid in the design of a Student/Machine software interface is discussed on the assumption that the student, the computer, and the software interface, can be considered as three independent, but interacting systems. By way of illustration a model is developed which could be used to design software for use in adult education. Chapter Five is in two parts, each part dealing with essentially the same concept - the transmission of knowledge about the process of educational software design. Two major strategies are considered. Firstly, the concept of a Microfactor is introduced as a way in which practitioners in the field of educational software design might communicate about solutions to certain problems. The chapter then proposes and discusses a unit of study for teachers on the topic of Educational Software Design in which practitioners communicate with beginners. The main focus of this unit, to be called &quoteEducational Software Design&quote, is on (1) Need for problem solving skills in educational Software Design; (2) Need for communication skills to facilitate collaboration between experts; (3) Need for a schema which will assist in the structuring of knowledge about educational software design. It is modelled on an existing unit in a BA(TAFE/ADULT) course which has been running for several years. A detailed description of this prototype unit and its design is given in appendix A and B. To conclude the study, Chapter Six considers some of the possible attitudinal barriers which can severely restrict the use of educational software. Even the most expertly designed software will be of no benefit if it is not used.