Ligand-based drug design (LBDD) is a powerful approach in the field of drug discovery that exploits the structural and physicochemical properties of known active molecules to identify and optimize new drug candidates. This chapter provides a comprehensive overview of LBDD, outlining its historical evolution from the earliest concepts of structure–activity relationships to the development of Quantitative Structure–Activity Relationships (QSAR) and pharmacophoric models employing the latest sophisticated techniques. After defining fundamental concepts such as the role of chemical descriptors in the representation of molecular structures and properties (0D, 1D, 2D, 3D, and 4D or SMILES and molecular graphs), the chapter focuses on significant advances in the integration of artificial intelligence and machine learning techniques to improve LBDD methods. Principles for building robust and reliable QSAR and pharmacophore models with AI are outlined, emphasizing the importance of curation of datasets, model development, and validation according to Organization of Economic Cooperation and Development (OECD) guidelines, a posteriori validations, and explainability. The chapter concludes by highlighting the potential impact of AI in modern drug discovery, presenting recent case studies, discussing limitations, and offering a perspective on future directions in this rapidly evolving field.