Access to clean energy has long been neglected within the humanitarian agenda. As a result, refugees are often locked into unsustainable and risky energy practices, especially in low-income countries. This study focuses on market-based interventions to deliver clean cooking fuels in refugee camps in protracted crises. The outcomes of market-based interventions are highly dependent on whether beneficiaries adopt and continuously use the products. However, clean cooking practices often face various adoption barriers. The aim of this study is to gain insights into the mechanisms and path-dependencies driving the adoption of clean cooking fuels in refugee camps and, by taking a modelling approach, to develop a method to analyse the effect of market-based clean cooking interventions. To this end, this study combines agent-based modelling with exploratory modelling techniques to capture social interactions, human decision-making behaviour, and deep uncertainties. The model formulation is grounded in a case study of a Rwandan refugee camp, and Diffusion of Innovation theory. Different interventions, including financial assistance for fuel purchase (cash transfers or vouchers), information campaigns, and maintenance activities, are simulated under a wide range of scenarios. Scenario discovery is applied to identify circumstances for success or failure of interventions. For both types of financial assistance, supporting information campaigns and maintenance activities are critical to create robust, timely, and long-term impact. Based on the findings, policy recommendations and guidelines for future interventions are derived.