How natural disasters affect conflict in war-torn countries is an important question, given the increasing prevalence of natural disasters due to climate change. Existing literature presents a disparate and seemingly contradictory findings on the disaster-conflict nexus, suggesting a need to examine potential moderators. This paper addresses this gap by introducing the presence of resource extraction sites as a key moderator for the relationship between natural disasters and violent incidents. I argue that third-party extractive actors mitigate post-disaster violence by influencing armed groups’ strategies, with the effect differing by resource type. In capital-intensive resource-rich territories (e.g., large-scale mining and oil and gas extraction), third-party actors stabilise post-disaster environment by buying off armed groups whose disruptive capacity is weakened by disasters. Conversely, in labour-intensive resource-rich territories (e.g., small-scale and artisanal mining), third-party extractive actors’ influence is limited and post-disaster disruption of the balance of power between armed groups leads to fights over the control of the resource. Using sub-national and geo-referenced data on violent incidents between the government and the rebels and resource licenses in Myanmar between 2010 and 2018, results show that while resource extraction sites generally mitigate the conflict-intensifying effect of natural disasters, this effect varies by resource type. The presence of capital-intensive resource extraction sites such large mining and oil and gas extraction are associated with lower risk of violent incidents compared to labour-intensive resource sites. Moreover, the operational timeframe of these extraction sites plays a crucial role: sites nearing license expiration display weaker mitigating effects on violent incidents.