The aim of this study was to explore and compare the results of occlusal load application to cusp-replacing composite restorations, studied by means of finite element (FE) analysis and in vitro load tests. A three-dimensional (3D) FE model was created with a set up similar to an in vitro load test that assessed the fatigue resistance of upper premolars with buccal cusp-replacing resin composite restorations. Occlusal load was applied to two geometries (with and without palatal cuspal coverage), and the tooth-restoration interface and composite material stresses were calculated. Subsequently, safety factors were calculated by dividing the material strength values by the obtained stresses. The highest safety factors were observed for the restorations with cuspal coverage. This was consistent with the load test, in which cuspal coverage led to higher fracture resistance. Furthermore, the FE analysis predicted that failure of the tooth-restoration interface is more likely than failure of the composite material. Correspondingly, the load test showed predominantly adhesive failures of the restorations. Although the described test methods did not lead to a complete understanding of the failure mechanism, it can be concluded that the FE analysis provides additional information with regard to the differences in fracture behaviour of these types of restorations.