In this study, social presence theory is proposed as an explanation for the relationship between permeability of boundaries and entitativity in online groups. This claim is tested in an experiment in which group boundaries are manipulated in a fictitious online discussion forum. Structural equation modeling and multiple group analysis were used to test the proposed model and determine whether results were different across the two experimental groups. Results indicate that although participants’ perceptions of boundaries were not different across the groups, boundaries predicts both social presence and entitativity. Further, boundaries may be better conceptualized as two separate constructs (weak and strong), and these constructs have differing, and unexpected relationships with entitativity and copresence. However, when additional established antecedents of entitativity are included in the model, boundaries no longer significantly predict entitativity. Implications of studying boundaries and copresence in online and takeaways for practitioners are discussed.