The increasing awareness that drugs may have the clinical effect through the interaction with multiple targets is encouraging the screening of investigational compounds across multiple biological endpoints. As the number and complexity of chemogenomics data sets increase, more computational approaches are being developed for the efficient analysis of structure-multiple activity relationships. In silico methods cover a wide range of applications including visual, qualitative, and quantitative approaches to describe in detail multiple ligand-protein relationships, find associations between targets and, whenever possible, to predict the bioactivity profile of small molecules. Here, we present a commentary of representative computational methods and their applications to characterize structure-multiple activity relationships and conduct the rational design of polypharmacology for the advancement of drug discovery.