During the development of new nonsteroidal anti-inflammatory agents, it was discovered that 1-aminoalkyl-3-aroylindoles have affinity for the cannabinoid brain (CB1) receptor. This has led to the development of over 100 cannabimimetic aminoalkylindoles, and the development of SAR for these compounds. Later work demonstrated that the aminoalkyl moiety was not necessary, and could be replaced by a four- to six-membered alkyl chain without loss of affinity. Investigation of these indoles led to the discovery of a CB2 selective ligand, 3- (1-naphthoyl)-N-propylindole. Subsequent work has provided several additional CB2 selective indoles. On the basis of a proposed pharmacophore for the cannabimimetic indoles, a series of pyrroles and indenes were developed, some of which are potent cannabinoids. SAR for several series of pyrroles have been developed. Two groups have described cannabimimetic indenes, which have been employed as rigid models for the receptor interactions of cannabimimetic indoles with the CB1 receptor. There is some evidence that the indoles bind to a somewhat different site on the receptor than traditional cannabinoids, and interact with the receptor primarily by aromatic stacking.