Thesis (M.Com. (Information Systems))--University of the Witwatersrand, Faculty of Commerce, Law and Management, School of Economic and Business Sciences, 2014. / Information systems discontinuance at the organisational level is an underrepresented research topic. However, it is an important problem in practice. The managerial decision to discontinue an existing information system by replacement, upgrade or abandonment requires extensive justification and evaluation. The timing of a system’s discontinuance is of strategic concern as the cost and impact of a packaged software application can be extensive. If the decision is made too early in the system’s life cycle, it can negatively affect the return on investment, but if the decision occurs too late, the system may become a liability to an organisation as it continues to consume resources in maintenance and support. A comprehensive framework is required that analyses the factors that determine discontinuance intentions of the decision makers within an organization. This study focuses on managerial perceptions of selected factors considered most important to discontinuance decisions. The study applies and extends the discontinuance framework proposed by Furneaux and Wade (2011) to develop a research model of the effects of these selected factors on organisational intentions to discontinue packaged software. Specifically, the research model hypothesizes the effects of ‘change’ forces that include the manager’s perceptions of system performance, system suitability and system supportability, as well as ‘inertia’ forces such as sunk cost, mimetic isomorphism and degree of the application’s integration into the enterprise architecture. The organisation’s ability to innovate, the age of the system and the presence of the initial sponsor are included as additional factors. To test the model, a survey methodology was employed to collect data from South African IT decision makers. The survey made use of a structured questionnaire instrument administered online. Useable data was collected from decision makers representing 103 organisations. Hypotheses were tested using regression and partial least squares structural modeling. Results showed that poor software performance, a lack of suitability and low levels of business and technical integration, were the strongest determinants of discontinuance intentions. The control effect of the software’s age relatively to the organisation’s portfolio of applications was proven to be significant in this research. This study has contributed by consolidating the effects of a number of factors drawn from the literature and developing an integrated framework of discontinuance. Results will help IT decision makers when considering a discontinuance decision by showing the relative effects of each factor and may be helpful to IT managers to determine the timing of the decision.