Fatty acid amide hydrolase (FAAH) is an enzyme responsible for the deactivating hydrolysis of fatty acid ethanolamide neuromodulators. FAAH inhibitors have gained considerable interest due to their possible application in the treatment of anxiety, inflammation, and pain. In the context of inhibitor design, the availability of reliable computational tools for predicting binding affinity is still a challenging task, and it is now well understood that empirical scoring functions have several limitations that in principle could be overcome by quantum mechanics. Herein, systematic ab initio analyses of FAAH interactions with a series of inhibitors belonging to the class of the N-alkylcarbamic acid aryl esters have been performed. In contrast to our earlier studies of other classes of enzyme-inhibitor complexes, reasonable correlation with experimental results required us to consider correlation effects along with electrostatic term. Therefore, the simplest comprehensive nonempirical model allowing for qualitative predictions of binding affinities for FAAH ligands consists of electrostatic multipole and second-order dispersion terms. Such a model has been validated against the relative stabilities of the benchmark S66 set of biomolecular complexes. As it does not involve parameters fitted to experimentally derived data, this model offers a unique opportunity for generally applicable inhibitor design and virtual screening.