G protein-coupled receptors (GPCRs) represent a significant class of drug targets, yet their diverse structure and complexity of ligand interactions present challenges for drug design. In this chapter, we focus on the use of ligand-based and structure-based approaches in GPCR drug discovery. Pharmacophore modeling is an important resource for determining critical ligand attributes and finding potential new GPCR drugs or improving lead compounds. The pharmacophore models, which can be generated  from crystal structure coordinates, homology modeling, or ligand-based methods, are utilized to design ligands with enhanced affinity, selectivity, and pharmacokinetic profiles. The central pharmacophore elements differ among GPCR classes, from ionic interactions in aminergic receptors to hydrophobic and H-bond motifs in peptide- and lipid-binding receptors. Ligand-based, structure-based, and hybrid strategies have been applied to computer-aided GPCR pharmacophore identification. QSAR modeling is another useful resource  for relating chemical structure to biological activity. Various 2D and 3D-QSAR approaches, such as CoMFA and CoMSIA, have been reported for several GPCR targets, whereas recent developments in machine learning support activity prediction and virtual screening. However, the concepts  of ligand bias and functional selectivity confuse classical QSAR and pharmacophore construction. Biased signaling is due to ligands that stabilize different conformations of the GPCR, which in turn preferentially activate distinct signaling pathways. The rational design of biased GPCR ligands, e.g., G protein-biased μ-opioid receptor agonists, must take all functional responses into account. The structure of GPCRs  in several states determined via X-ray crystallography and cryo-EM reveals possible mechanisms by which ligands are recognized and activate the receptors. This allows for a functionally targeted approach to structure-based drug design, one directed toward distinct classes of conformational states. The integration of experimental and computational methods to study GPCR dynamics and polypharmacology will  further promote the development of safer and more effective drugs.