In the last two decades the software reliability research is becoming more accessible and popular. Important place for such research is taken up for enterprise resource planning (ERP) systems as the most important software systems for different businesses. The complexity of their architecture and supported business processes is resulting with complex and demanding reliability analysis. Existing studies generally examine the impact of operational risk in the field of hardware, information security, communications and business processes but less in the field of software errors, training, configuration and customizations. Customers need simple and easily understood procedures in assessing the reliability of the ERP system to act before the system becomes unusable or at worst undermine their business. In the introduction of this dissertation are presented the basic concepts and terms related to ERP systems with the concept of defect tracking in their implementation and production. The ERP system life cycle processes are particularly described with emphasizing of adjustments and adaptations as sources of potential defects. The basic definitions and software reliability models are presented with particularly description of later used exponential and Weibull model of software reliability. Later on, the process of making an appropriate reliability model from the available measurements of the real ERP system is described. The procedures of processing and analyzing the collected data for easier later identification of the appropriate statistical distributions are highlighted. Briefly, presentation of software package used for later statistical fitting of empirical data is made with an example from performed case study. Mathematical algorithms that can be used for Weibull and exponential software reliability models in assessing the best-fit statistical distribution of empirical data are specifically mentioned and explained. The statistical tests that were used and the way of their application in evaluating the confidence of the chosen ERP reliability model are presented. The operational profile technique is described as an interesting method for efficiently ERP system testing and reliability growth. Theoretical concepts are systematically processed on a specific case study of specific ERP software modules and in an understandable manner presented with illustrations how the operating profile should be developed and applied. Constant upgrading of ERP systems is necessary, but can cause new defects. This dissertation attempts to model the likelihood of defects after completed upgrades with Weibull defect probability density function. A case study is presented analyzing data of recorded defects obtained for one ERP subsystem. The trends are observed for the value of the parameters relevant to the proposed statistical Weibull distribution for a given one year period. As a result, the ability to predict the appearance of defects after the next upgrade is described. The result of the applied modelling is in modelling the reliability of the ERP system from a user perspective with estimated parameters like expected maximum number of defects in one day or predicted minimum of defects between two upgrades. The results of this thesis will be useful to manufacturers of ERP systems in the modelling of the quality of implementation of their ERP products, but also for ERP users in assessing the reliability of the applied ERP system.