The first part of the paper examines Structure-Activity Relations (SARs) and their components from a very general point of view. The various types of interpretation emerging from statistically valid relations will be examined, namely causal (mechanistic), contextual (empirical), fortuitous, and tautological correlations. Implications for ADME predictions will be seen when discussing the diversity of interactions between active compounds (e.g., drugs) and biological systems. The second part of the paper is more specific and presents the concept of molecular-property space, an all but neglected concept in SARs. Recent results from Molecular Dynamics (MD) simulations and Molecular Interaction Fields (MIF) computations of acetylcholine will be used to illustrate not only the well-known conformational space of this molecule, but also its property space as exemplified by its lipophilicity space. It will be seen that a molecule as small as acetylcholine is able to span a relatively broad property space. Most significantly in an ADME perspective, the molecule is able, within the limits of its property space, to adapt to the medium. This is equivalent to saying that the medium constrains the molecule to resemble it as much as feasible.