Introduction The possible link between the molecular structure and a biological activity was pointed out by Linus Pauling (1946) just before the middle of the 20th century. Nevertheless, this concept has been already suggested a century before by Emil Fisher and “lock-and-key” theory: “To use a metaphor, I would say that enzyme and glucoside are such lock and key that fit together in order to exert a reciprocal chemical effect on each other”a (Fischer, 1894). The notion that the biological properties of a compound would be a function of its chemical characteristics is now well admitted, and is closely linked to that of pharmacophore. The pharmacophore concept is related to the interaction of drugs, and more broadly to biological molecules, with specific receptors. Such interaction causes a functional response that, therefore, produces an observed biological activity. The work of Paul Ehrlich (1897) is usually attached to the notion of pharmacophore (Selassie, 2003); nevertheless, it seems to be controversial because the term “pharmacophore” does not appear in Ehrlich’s articles (Van Drie, 2007; Guner and Bowen, 2014). Currently, the characteristics which define a pharmacophore are provided by the following IUPAC definition (Wermuth et al., 1998): “A pharmacophore is the ensemble of steric and electronic features that is necessary to ensure the optimal supramolecular interactions with a specific biological target structure and to trigger (or to block) its biological response.” The development of the computational modern concept of quantitative structure– activity relationships (QSAR) occurred in the second half of the 20th century due to the improvement of computing capacity, and to the work of Hansch and Fujita (1964). Inclusion of tridimensional (3D) molecular characteristics led to the more recent 3D-QSAR methods (Livingstone, 2000; Martin, 1998; Perkins et al., 2003; Yang and Huang, 2006). Since about 10 years, all approaches based on the molecular similarity principle, which specifies that similar molecules or chemical compounds tend to have similar properties (Martin et al., 2002; Bender and Glen, 2004; Brown, 2009; Khanna and Ranganathan, 2011; Maggiora et al., 2014) are grouped into the chemoinformatics field (Gasteiger, 2006a; Agrafiotis et al., 2007; Brown, 2009; Varnek and Baskin, 2011; Willett, 2011; Vogt and Bajorath, 2012); however, QSAR/QSPR terminology is still used (Cherkasov et al., 2014). Nowadays, the applications of the QSAR approach are major in drug design (Yang and Huang, 2006; Gasteiger, 2014). Nevertheless, QSAR studies are also applied to the prediction of physicochemical properties, in which H. Wiener (1947) has been the pioneer; such studies are commonly termed as quantitative structure–property relationships (QSPR) (Katritzky et al., 1995). In fact, there are many areas involved in QSPR studies. A large applied domain concerns the prediction of retention coefficient in chromatography (Heberger, 2007; Beteringhe et al., 2008), and some of these studies are called quantitative structure–retention relationships (QSRR). QSAR/QSPR approaches are now also largely used in (eco)toxicology (REACH) (Lahl and GundertRemy, 2008; Nicolotti et al., 2014). In silico prediction of aqueous solubility and log P constitutes, for many years and still currently (Jorgensen and Duffy, 2002), a major challenge, aptly summarized by the following article title: “Why are some properties more difficult to predict than others? A study of QSPR models of solubility, melting point, and Log P” (Hughes et al., 2008), and a lot of works are devoted to this issue (Dearden, 2006; Llinas et al., 2008; Hewitt et al., 2009; Hopfinger et al., 2009; Wang and Hou, 2011; Ali et al., 2012; Chevillard et al., 2012; Raevsky et al., 2015). Some of these works are applied to odorant molecules, to studies of both structure– activity, and structure–property relationships. They are presented in the second part of this chapter, after a first part devoted to the presentation of chemoinformatics and the QSAR concept.