Benzodiazepine (BZ) anxiolytics mediate their clinical effects by enhancing the effect of gamma-aminobutyric acid (GABA) at the GABA-A receptor. Classical BZ full agonists such as diazepam, which maximally enhance the function of GABA-A receptors, are effective anxiolytics but carry unwanted side effects including sedation, dependence and abuse liability, limiting their utility. Although a second generation of 'partial agonist' BZs have been pursued, promising preclinical data, in terms of anxiolytic efficacy and decreased unwanted effects, have so far failed to translate to the clinic. Following the insights into GABA-A receptor subtypes mediating the effects of BZs, a third generation of 'receptor subtype-selective' BZ site ligands have been developed. However, it remains to be determined whether promising preclinical data are recapitulated in the clinic.