Militant organizations commonly break down and split apart, with new groups emerging from the ranks of existing organizations. From Syria to Iraq to Afghanistan, militant groups have splintered and proliferated in this way, creating fragmented oppositions that significantly complicate the conflict landscape. This process of organizational splintering historically has created some of the deadliest and most well known organizations including Al Shabaab, Black September, and the Real IRA. However, at other times the new organizations have quickly disappeared, failing to impact the conflict in any meaningful way.What explains this variation in the trajectory of militant splinter groups over time? Specifically, this dissertation explores why some organizational fractures produce new groups that are durable and increasingly radicalized, while others merely fall apart. This is an important topic that has ramifications for how academics and policymakers alike understand the behavior of specific actors and also the evolution of fragmented conflicts around the globe. I develop a new theory to explain variation in rates of survival and radicalization that focuses on the content and the consistency of internal organizational preferences. I argue that the content of group preferences can explain relative rates of radicalization and tactical change whereas the consistency or alignment of those preferences influences their chances of survival. Splinter groups that attract tactical and strategic hardliners are most likely to radicalize while inconsistent internal preferences lead to feuding, a lack of cohesion, and a lower likelihood of survival. Although impossible to directly observe, I show that different pathways of organizational breakdown, which one can observe, strongly shape the distribution of group preferences. In other words, different pathways of group formation have enduring effects on organizational behavior. I test my theory with a mixed-methods research design. The empirical results from analyzing a new data set provide robust cross-national support for my theory while my case study of republican militants in Northern Ireland - supplemented by three months of field world in Belfast, London, and Dublin - demonstrates the theory's causal mechanisms in action. These findings confirm that the conditions leading to group formation play an enduring role, driving group behavior well into the future.