Experiments were conducted to examine the molecular basis for the high affinity and potency of a new class of 5-HT(2A) receptor agonists, N-benzyl phenethylamines. Competition binding assays at several serotonin receptors confirmed that an N-arylmethyl substitution was necessary for affinity increases up to 300-fold over simple N-alkyl homologs, as well as enhanced selectivity for 5-HT(2A) versus 5-HT(2C) and 5-HT(1A) receptors. PI hydrolysis functional assays confirmed that these N-benzyl phenethylamines are potent and highly efficacious agonists at the rat 5-HT(2A) receptor. Virtual docking of these compounds into a human 5-HT(2A) receptor homology model indicated that the N-benzyl moiety might be interacting with Phe339((6.51)), whereas the phenethylamine portion was likely to be interacting with Phe340((6.52)). Experiments in h5-HT(2A) receptors with Phe339((6.51))L and Phe340((6.52))L mutations seem to support this hypothesis. Dramatic detrimental effects on affinity, potency, and intrinsic activity were observed with the Phe339((6.51))L mutation for all N-benzyl analogs, whereas most N-unsubstituted phenethylamines and traditional agonists were only weakly affected, if at all. Consistent with other published studies, the Phe340((6.52))L mutation detrimentally affected affinity, potency, and intrinsic activity of nearly all compounds tested, although a strong change in intrinsic activity was not seen with most N-aryl analogs. These data further validate the topology of our h5-HT(2A) receptor homology model. It is noteworthy that this study is the first to identify a hitherto unrecognized role for residue 6.51 in agonist activation of a serotonin G protein-coupled receptor (GPCR), whereas most previous reports have suggested a varied and sometimes contradictory role in homologous GPCRs.