Complementary 3D-QSAR modeling of binding affinity and functional potency is proposed as a tool to pinpoint the molecular features of the ligands, and the corresponding amino acids in the receptor, responsible for high affinity binding vs those driving agonist behavior and receptor activation. This approach proved successful on a series of nicotinic alpha(4)beta(2) ligands, whose partial/full agonist profile could be linked to the size of the scaffold as well as to the nature of the substituents.