A novel concept, the fracture efficiency, is proposed in this dissertation as a performance index of fracture specimen designs for adhesive bonds. Evaluated based on a simple quantity, the fracture efficiency parameter, the fracture efficiency represents how much crack driving force a certain specimen design can produce along the bondline for a given maximum non-singular stress in the adherends. In this dissertation, a study of membrane peeling specimens shows that various existing test geometries have little difference in the fracture efficiency. Furthermore, the study shows that it is unlikely to develop new n1embrane peeling tests with high fracture efficiency. Therefore, gross inelastic energy dissipation in the film specimen may occur when using the membrane peeling tests, and may result in large errors in the film adhesion measurement when only elastic material response is considered in the analysis. As a result, the use of a backing material for film adhesion measurement is suggested because of the significant increase in the fracture efficiency when a proper backing material is used. Besides the use as one of the guidelines for fracture specimen designs, it is also demonstrated in this dissertation that the fracture efficiency parameter can also be used to help one determine the validity of a measured bond fracture strength. The study of fracture efficiency leads to new analyses of several existing fracture test geometries, and contributes to a better understanding of the standard blister, island blister, peninsula blister, peel, and cracked lap shear tests. It is found that the island blister and peninsula blister tests can produce very high energy release rates at low pressures by using a small island radius or a small peninsula width. However, these so called "high energy release rate" tests also induce localized high stress concentrations near the debond front and therefore, yielding or rupturing may still unavoidable in such specimens. The significant localized inelastic effect on the measurement of the bond fracture strength is also demonstrated in the peninsula blister experiment for an adhesive tape specimen subjected to a relatively low pressure load. Another important conclusion drawn from the analyses of various membrane peeling tests is that the existing membrane blister tests are special cases of the peel test if compared very near the debond front. Besides analyses for membrane peeling specimens, this dissertation also reports a new closed form solution for the cracked lap shear specimen using a geometrically nonlinear beam-column approach. Excellent agreement for energy release rates has been shown between the closed form solution and geometrically nonlinear finite element analyses. The study also discusses how to design a cracked lap shear specimen of constant energy release rate and high fracture efficiency. / Ph. D.