High values for the work of separation have been reported in peel tests on fibrillating interfacial systems. The exact origin of these high values is not properly understood, since it remains unclear which dissipative mechanisms related to fibrillation cause a significant increase in the work of separation. In this paper, the contribution of fibril mechanics to the work of separation is quantified. To this end, a micromechanical model of a single fibril is used, in which the growth of a nucleated fibril up to the moment of fracture is described. The initial geometry is varied to reflect the variability in the substrate profile. It is observed that the stresses and strains that occur in the model are well beyond typical bulk values. Given the large variation in measured stress-strain curves for rubber materials reported in literature, a small scale single fibril experiment is performed to assess the applicability of reported bulk material parameters to this problem. The obtained experimental response falls well within the model bandwidth for the range of material parameters from literature. Furthermore, the fibril fracture stress, which serves as the failure criterion in the model, is extracted from the experiment. From the micromechanical model results, it appears that, for the considered range of material properties, the work of separation is mainly influenced by the fracture stress. Furthermore the initial geometry has a profound influence on the obtained work of separation. It is concluded that the work of separation determined from the micro-model is significantly larger than the intrinsic adhesion energy, yet it remains still an order of magnitude smaller than the values reported in literature. For the considered system this indicates that, although it has a significant contribution, fibril deformation is not the only contribution to the high work of separation.