This thesis investigates the influence of specimen width on the mode-I fracture toughness of adhesively bonded plate-like specimens. First, the influence of different double cantilever beam (DCB) loading arrangements on the measured fracture toughness is investigated and an accurate method for determination of the traction-separation law of the interface using the direct approach with the digital image correlation (DIC) is presented. An additional set of DCB experiments with three different specimen widths is conducted to study the influence of the width on the measured fracture toughness. A novel technique for measuring the relative displacement of the plates with full-field DIC, termed Symmetry Based Top Surface Analysis (SBTSA), is developed to measure the actual crack position and shape, as well as the crack area increments. By using this value in the finite-difference form of the definition of critical energy release rate (Area method) any dependence on the beam theory assumptions is avoided. Experimental results suggest that wider specimens have substantially higher fracture toughness. An original numerical model was developed using the Q4U3 plate elements connected by the kinematically compatible interface elements with an embedded exponential traction-separation law. A numerical study of the efficiency and robustness indicates good convergence properties of the proposed model. The aforementioned approach to compute the fracture toughness using SBTSA and the Area method was validated by means of virtual experiments. The comparison of experimental and numerical results for DCB specimens with different widths shows good agreement of the global load-displacement response, with an excellent agreement of the crack front shape and size of the damage process zone.