This thesis attempts to explain why some domestic crises escalate to internationalized civil war, while others do not. Existing research on unilateral intervention in civil war does not pay sufficient attention to the dyadic nature of conflict, and the decision by an actor to resist intervention. Jack Snyder’s (1991) theory of Great Power “regime cartelization” is here adapted to explain why some transitional regimes in small states are less susceptible to immediate, extended deterrence by external actors that support separatist domestic challengers. Cartelized regimes are transitional regimes with relatively weak democratic institutions, and executive decision making influenced by nationalist ideology. The main claim of this thesis is that regime cartelization is positively related to the onset of internationalized civil war, given that secondary party support to the domestic challenger is staunch. This is because cartelized regimes prefer the cost of war over the audience cost of backing down from a contestation. A comparative, qualitative case study of two domestic crises in Georgia 2004-2008, and two domestic crises in Ukraine 2014-2016 supports this claim.