TPA023, a GABA(A) alpha2,3 alphasubtype-selective partial agonist, is expected to have comparable anxiolytic efficacy as benzodiazepines with reduced sedating effects. The compound lacks efficacy at the alpha1 subtype, which is believed to mediate these effects. This study investigated the effects of 0.5 and 1.5 mg TPA023 and compared them with placebo and lorazepam 2 mg (therapeutic anxiolytic dose). Twelve healthy male volunteers participated in this placebo-controlled, double-blind, double-dummy, four-way, cross-over study. Saccadic eye movements and visual analogue scales (VAS) were used to assess the sedative properties of TPA023. The effects on posturaL stability and cognition were assessed using body sway and a standardized battery of neurophysiological memory tests. Lorazepam caused a significant reduction in saccadic peak velocity, the VAS alertness score and impairment of memory and body sway. TPA023 had significant dose dependent effects on saccadic peak velocity (85 deg/sec maximum reduction at the higher dose) that approximated the effects of lorazepam. In contrast to lorazepam, TPA023 had no detectabLe effects on saccadic latency or inaccuracy. Also unlike lorazepam, TPA023 did not affect VAS alertness, memory or body sway. These results show that the effect profile of TPA023 differs markedly from that of lorazepam, at doses that were equipotent with regard to effects on saccadic peak veLocity. Contrary to lorazepam, TPA023 caused no detectable memory impairment or postural imbalance. These differences reflect the selectivity of TPA023 for different GABA(A) receptor subtypes.