Decision support systems (DSS) provide appropriate information for making effective decisions relevant to business needs. DSS development technologies however, have shown limited success in many problem areas, including rural business applications. DSS developed for rural business decision makers have been limited by a range of issues. These include low user uptake, retaining rigid analyses in the face of changing industry needs, and the need for ongoing reengineering effort to incorporate new information in the current system. User factors such as a limited fit with their specific context of decision making problems, differences in problem definition between users and designers, and poor usability, generally only compound this situation. Such issues have led to the low adoption rate of DSS within the rural sector. A review of the available literature reveals a significant gap in finding technical solutions for these problems. While there are several approaches to developing DSS including expert systems, model-based systems, hybrid systems and user-developed systems, each has limitations. Expert systems do not support fully functional DSS development and their well-known limitations include obsolescence and lack of context sensitivity. Whilst model-based and hybrid decision systems may overcome some limitations, they do not represent knowledge in forms that typical end users can understand. On the other hand, while end-user developed applications reportedly achieve higher user uptake, they suffer from local subjectivity, quality and business alignment issues. End-user development is also often associated with risks including data disarticulation, threats to data integrity and security, and the inability to interoperate with other systems. From a practical viewpoint, current DSS development approaches also fail to accommodate rapidly changing factors necessary for effective decision making in rural businesses. These factors include external markets regulation and legislation, and new scientific discoveries. Despite significant need for decision support, no prior research has satisfactorily addressed these problems.