All living organisms, including human beings, are able to detect and discriminate myriads of structurally diverse odorants through their interaction with olfactory receptors (ORs) (1). It is well accepted that the perception of thousands of odors by about 380 ORs results from a combinatorial coding, in which one OR recognizes multiple odorants and different odorants are recognized by different combinations of ORs (2). In a previous study (3), the functional characterization on two human ORs, called OR1G1 (class II) and OR52D1 (class I) have been performed using 95 odorant molecules. We used these previously obtained functional data (3) to perform a molecular modelling study of ligands using Catalyst/HypoGen software (Catalyst version 4.11, Accelrys Inc., San Diego, 2004, running on PC Red Hat Enterprise Linux). Catalyst/HypoGen takes into account molecular flexibility by considering each compound as a collection of conformers. It generates models, named “hypotheses”, which describe ligands as sets of chemical functions. These hypotheses should be able to predict the activities of different compounds having the same receptor binding mechanism. In a previous work we obtained an alignment model of OR1G1 ligands, which satisfactorily explained the experimental activities, and permitted to predict novel agonists for OR1G1 which were experimentally validated (4). In the present work, we applied the same procedure to activity data of OR52D1 ligands and compare the best significant hypotheses models obtained for both OR1G1 and OR52D1 ligands. We attempted to decipher the odotopes of OR1G1 and OR52D1 agonists in order to investigate the role of ORs in olfactory coding.