The literature on rebel cohesion/fragmentation currently presents valid macro-level explanations for rebel group cohesion/fragmentation, yet no model currently exists exploring non-utility-based motivations regulating pro-group behaviour. This thesis tests the assumptions of a novel social psychological framework – Identity Fusion Theory – on rebel cohesion. Using a primarily quantitative approach applying a logistic regression model to primary data gathered on the Syrian Civil War, and a smaller qualitative element comparing groups with “fused” and “non-fused” memberships during the war, the thesis tests the hypothesis that “groups with highly fused memberships will be less likely to fragment during the course of a civil war than groups with non-fused memberships”. The results of the analysis show a strong, positive relationship between fused memberships and rebel cohesion. The results of the thesis are to some extent limited by potential omitted variable bias (suggested by high R2 values) and the use of Syria as the only population from which to sample. Nevertheless, the relatively large number of observations in the dataset (63) as well as heterogenous nature of groups involved suggest the findings are generalisable. In conclusion, we did find a positive relationship between fusion and rebel group cohesion. The main implications of this thesis being that future academic research may benefit from focusing on social psychological factors when examining rebel dynamics while policymakers potentially shape better responses to insurgencies and rebellions.