The interface toughness of adhesively bonded structural members is one of the critical parameters for adhesive joint design. It is often assumed that the joint toughness is a material constant so that its value can be obtained from fracture tests of simple geometries such as DCB for Mode-Ⅰ, ENF for Mode-Ⅱ, using linear elastic fracture mechanics (LEFM). However, the LEFM assumption of point-wise crack-tip fracture process is overly simplistic and may cause significant error in interpreting fracture test data. In this paper, the accuracy and applicability of various traditional beam-bending-theory based methods for fracture toughness evaluation, such as simple beam theory (SBT), corrected beam theory (CBT) and experimental compliance method (ECM), were assessed using the cohesive zone modelling (CZM) approach. It was demonstrated that the fracture process zone (FPZ) size has profound influence on toughness calculation and unfortunately, all the classic beam-bending theories based methods fail to include this important element and are erroneous especially when the ratio of crack length to FPZ size is relatively small (＜5.0). It has also been demonstrated that after the FPZ size is incorporated into simple beam formulations, they provide much improved evaluation for fracture toughness. Formulation of first order estimate of FPZ size is also given in this paper.