With the decline of conventional gas resources in many industrialised regions, energy security has increasingly become a topic of interest. Fuelled by disturbed diplomatic relations, both national and regional governments have the desire to decrease their dependency on foreign nations for their gas supply. However, the energy future of these regions is uncertain, and energy markets are highly complex and dynamic. The use of simulation modelling might provide a solution for discovering system behaviour of complex multi-actor systems, such as the gas market. Over the past decades, multiple simulation methods have evolved, including System Dynamics (SD) and agent-based modelling (ABM). Although both methods have been applied to overlapping areas of interest, they both are fundamentally different in terms of conceptualisation, formalisation and implementation, which is likely to impact the outcomes of the simulation exercise. In this research, a multi-method approach is proposed as a way to explore the effects of method uncertainty and to improve simulation modelling for policy analysis by tapping into the relative strengths of both SD and ABM. The question that is central to this research is: What are the merits of a multi-method approach in scenario discovery? An answer will be provided to this question by applying the multi-method approach to the case of regional import dependency for natural gas. For this case, the aim is to discover and explore import dependency scenarios for selected world regions, given the relevant uncertainties that are affecting these scenarios. In specific, the effects of global shale gas development will be explored. Two models, one SD model and one AB model, are provided with partially identical inputs, and will be treated as equally ‘true’ representations of the real-world system. With regard to the case, it has been shown that the gas market and thus import dependency is strongly demand-driven, though also other leverage points exist to steer the system towards lower levels of import dependency. For the Far East region, incentives are required to push domestic gas production towards higher levels. Depending on the assumptions made regarding the economic growth of regions and the size of their resources, all regions may face both high (>50%) and low (<10%) import dependency. The logical next step would be to define regional policies and test their robustness and performance across both models. However, the current models are limited in the uncertainty spaces that are explored, and, depending on the type of policy, have to be adapted to allow for the testing of policies. With regard to the multi-method approach, it has been shown that SD models and AB models are able to complement each other in terms of capturing system behaviour. The multi-method approach has led to the discovery of system behaviour that would otherwise have remained unknown. The implications for policymaking are twofold; on the one hand a multi-method approach provides a stepping stone towards more robust decision-making, on the other hand it poses an additional challenge for finding, defining and evaluating the performance of policies. More experience with the multi-method approach in various types of problems is required to provide a more decisive answer to the main research question.