There is a growing threat from terrorism and insurgencies worldwide in recent years. It is puzzling why some initially weak militant groups, who face immense difficulties in garnering material resources and support, sustain violent operations and confront more powerful militaries. Why do some militant groups engage in sustained armed conflicts while other groups do not? Understanding this phenomenon is critical since groups that sustain military operations gain more influence, recruitment, and fundraising capabilities while further weakening the target state. Relying on a resource mobilization framework, a quantitative regression analysis of 246 prominent militant groups identified in the Global Terrorism Database (GTD) from 1970-2007 finds that organizational characteristics, on average, are better predictors of sustained armed conflict than measures of group capabilities – diverging from current explanations of insurgency onset or outcomes. Posing a serious challenge to a state is not necessarily a function of how powerful or capable a group may seem – it’s more about the competitive militant environment and internal capacity to effectively mobilize resources and maintain armed hostilities against regime forces. More specific findings suggest that religious militant groups, which rarely achieve ultimate objectives, are more likely to engage in sustained armed conflicts than other ideologically oriented groups. Further challenging conventional wisdom, groups with relatively less centralized command and control are as likely to engage in sustained armed conflicts as the most hierarchically structured organizations. More competitive militant environments also reduce the likelihood any particular group presents a major threat to the states they fight. This study finds that the vast majority of militant groups that engaged in sustained armed conflicts were the most dominant group in their environment around the time the group challenged the state. In the nascent stages of an insurrection, militant groups often have to consolidate rivals to dominate a constituency before taking on the regime. These insights offer important policy implications, considering it is far more difficult for states to defeat a full-fledged insurgency than prevent a nascent insurrection from flourishing. Acknowledging the limits of large-n research, this paper presents generalizable empirical associations across diverse militant groups and helps identify key cases for in-depth analysis by the author in subsequent work.