Structural similarity is one of the basic underlying principles in drug discovery and development. Numerous algorithms and concepts are known to search compound libraries for analogous compounds assuming that similar compounds show similar biological activity. In recent years the focus shifted towards more complex methods using 3D-shape similarities on one side and highly reductionistic approaches such as smiles substrings on the other side. Furthermore, pharmacological activity profiling becomes increasingly important. Within this review we will highlight selected new concepts and methods applying similarity metrics in the drug discovery and development process.