G Protein–Coupled Receptors are the largest family of membrane proteins with over 800 members expressed in humans. Class A (rhodopsin-like) GPCRs is the largest and the best known subfamily with numerous important pharmacological targets (playing a key role, e.g., in the etiology and treatment of the nervous system diseases). Especially interesting for medicinal chemistry are serotonin receptors. This thesis presents an in silico methodologies improving the search for new serotonin receptor ligands, covering four areas of computer-aided drug design: pharmacophore mapping, homology modeling, bioisosteric replacement and fingerprint reduction.Computational methods are undeniably an essential phase in the process of designing the new drugs. The methods described in this thesis create not only new perspectives in the search for the new serotonin receptors ligands, but can be expanded to any other biological target.