This work developed a discrete element model to adaptively capture the transitional micromechanical response and failure mode of adhesive joints with dissimilar adherend materials and different configurations. Especially, the development of the model only requires one-time calibration. Loctite EA 9497 epoxy adhesive, aluminium (AL) and polyphthalamide (PPA) were selected to make different types of hybrid adhesive joints for the lab tests. The modelling applicability in simulating Mode I, Mode II cohesive failure and adhesive, mixed failure modes was subsequently validated with the experimental data. The validation shows that the proposed model can accurately capture the observed failure modes and joint performances. It is followed by investigating the ability to adaptively obtain the variation of fracture energies with different adhesive thicknesses. The results agree well with the current reports on the growth trend of fracture energies which see a rise till the thickness reaching 0.8 mm and subsequently decline to a plateau. Finally, key factors including the adhesive thickness, lap length were selected to perform a parametric study to investigate their influences on the failure mechanism and micromechanical response of hybrid joints. It is found that a thickness range of 0.1–0.3 mm is adequate to obtain satisfactory joint strength whilst thicker adhesive over 0.6 mm will decrease the joint strength. This is due to that a thinner adhesive layer can facilitate its cohesive fractures and thus fully use the resistance of adhesive. A range of lap length from 6 mm to 12.5 mm was found to have a higher efficiency of improving the joint strength when AL adherend was used.