All substantive areas of law with no exception have a common concern for the processes by which legal disputes get resolved. Naturally, the success of any particular litigation strategy in a legal dispute depends on a number of factors. Examples of influential factors are procedural costs, the judge’s accuracy and, most importantly, the litigation strategy followed by the counterpart in dispute. Previous work has focused on analysing legal disputes individually, thus providing an answer to the question: What litigation strategy may be most appropriate when confronted with a particular counterpart’s strategy? The problem, of course, is that the counterpart’s strategy is rarely known in advance. In contrast, in this paper we adopt a dynamic view of the legal system as a whole. To do this, we assume that the most successful litigation strategies at a certain time are more likely to be followed in the future, so the prevalence of different strategies in the system will generally change in time. Importantly, this change in the frequency of litigation strategies in the legal system will in turn affect the relative success of each litigation strategy, thus creating a double feedback loop between prevalence and success of litigation strategies, which we aim to explore in this paper. Thus, the subject main purpose of this paper is to offer a novel approach to study legal disputes, looking at the whole litigation system as a single entity that evolves through time. In particular, we focus on cases of medical liability, and use agent-based simulation to provide a dynamic view of how various factors, such as the magnitude of legal expenses and the accuracy of the judicial system, affect the type of litigation strategies that are successful and prevail in a certain judicial context.