Although the relationship between religion and violent internal conflict is increasingly studied in the civil war literature, previous studies largely focus on factors influencing the onset of armed conflict. This thesis examines the less analyzed aspects of conflict intensity and duration. More specifically, it examines how these aspects are influenced by the presence of identity-based religious cleavages. By applying a theoretical perspective novel to the religion-conflict nexus, the thesis seeks to provide theoretical knowledge on how faith affects conflict dynamics. Concerning intensity, it is argued that religion, as a basis for identity and organized around a common belief-system and common doctrine, relaxes intragroup problems and makes it easier for belligerents to mobilize. Regarding duration, it is proposed that religious cleavages make it harder for the parties to establish the intergroup trust needed to reach stable peace agreements. Through extensive data collection a new indicator is introduced, measuring the presence of identity-based religious cleavages in 241 intrastate conflicts in the period 1946-2004. Results show that religious conflicts, as defined, are significantly more intense than non-religious ones. Furthermore, the analysis reveals an ambiguous impact on duration. In early stages religious conflicts are more likely than others to be terminated, whereas conflicts that have lasted at least two and half years are less likely to be terminated if they involve a religious cleavage.