The main objective of the work has been to develop theoretical methods that allow prediction of the strength of different adhesively bonded joints, and to evaluate the methods by comparing predictions with results achieved from testing of real joints. The theoretical methods include both analytical and numerical based methods. The numerical calculations were carried out with the finite element program ANSYS. The Matlab software was helpful in deriving analytically based results. The work consists of two main parts. First, double-lap joints between steel plates are studied. Different analytical methods are used to predict the shear stress, which is the dominant stress in these joints. These predictions compared very well with similar results obtained by numerical analyses. This first part was carried out primarily for the purpose of gaining knowledge and insight needed to carry out the second part work. In part two, joints between steel sandwich plates were studied. The objective of this part was mainly focused on strength predictions, and on the effect on such predictions of not including, or including, non-linear effects like yielding. Comparisons of test results and predictions, that were not as good as was hoped for, are made and discussed further in the report.