Objectives: To investigate the influence of different resin composite and glass ionomer cement material combinations in a "bi-layer" versus a "single-layer" adhesive technique for class I cavity restorations in molars using numerical finite element analysis (FEA). Materials and Methods: Three virtual restored lower molar models with class I cavities 4 mm deep were created from a sound molar CAD model. A combination of an adhesive and flowable composite with bulk fill composite (model A), of a glass ionomer cement with bulk fill composite (model B) and of an adhesive with bulk fill composite (model C), were considered. Starting from CAD models, 3D-finite element (FE) models were created and analyzed. Solid food was modeled on the occlusal surface and slide-type contact elements were used between tooth surface and food. Polymerization shrinkage was simulated for the composite materials. Physiological masticatory loads were applied to these systems combined with shrinkage. Static linear analyses were carried out. The maximum normal stress criterion was adopted as a measure of potential damage. Results: All models exhibited high stresses principally located along the tooth tissues-restoration interfaces. All models showed a similar stress trend along enamel-restoration interface, where stresses up to 22 MPa and 19 MPa was recorded in the enamel and restoration, respectively. A and C models showed a similar stress trend along the dentin-restoration interface with a lower stress level in model A, where stresses up to 11.5 MPa and 7.5 MPa were recorded in the dentin and restoration, respectively, whereas stresses of 17 MPa and 9 MPa were detected for model C. In contrast to A and C models, the model B showed a reduced stress level in dentin, in the lower restoration layer and no stress on the cavity floor. Significance. FE analysis supported the positive effect of a "bi-layer" restorative technique in a 4 mm deep class I cavities in lower molars versus "single-layer" bulk fill composite technique.