In multiparty civil wars, why do some rebels form alliances with one another but some do not; while still others divert resources to inter-rebel conflict? Although there are several answers to this question, most do not consider the role of rebel-public ties. Yet a group’s interactions with the public matters, as research have shown that constructive ties with the public enhance a group’s relations with the government, third-parties, along with its fighting effectiveness and endurance. This thesis expects rebels with more constructive ties to the public to be more selective partners and more rivalrous against other groups. However, groups with destructive ties would be less selective and rivalrous. Employing logistic regressions over cross-sectional time-series datasets, this thesis finds that former types of groups are more likely to engage in inter-rebel conflict to eliminate their rivals. Yet when the state becomes a non-credible actor to bargain with, where both gaining a concession and co-optation of other groups by the state are unlikely, such rebels tend to avoid conflict in an attempt to form a united body of opposition peacefully. Contrarily, those having destructive ties with the public avoid conflict with others due to the numerous concomitant deficiencies they have. Finally, groups do not differ from others in their alliance behavior based on their ties to the public. The findings imply that public support is considered mainly when rebels decide whether to make limited resources scarcer through engaging in inter-rebel conflict where they may need a last resort.