The use of non-selective gamma-aminobutyric acid (GABA) enhancers, such as benzodiazepines in the treatment of anxiety disorders is still widespread but hampered by unfavourable side effects. some of these may be associated with binding properties to certain subtypes of the GABA(A) receptor that are unnecessary for therapeutic effects. MK-0343 was designed to be a less sedating anxiolytic, based on reduced efficacy at the alpha1 subtype and significant efficacy at alpha2 and alpha3 subtypes of the GABA(A) receptor. This paper is a double-blind, four-way cross-over (n = 12) study to investigate the effects of MK-0343 (0.25 and 0.75 mg) in comparison to placebo and an anxiolytic dose (2 mg) of the non-selective agonist lorazepam. Effects were measured by eye movements, body sway, Visual Analogue scales (VAS) and memory tests. Lorazepam impaired saccadic peak velocity (SPV), VAs alertness scores, postural stability and memory and increased saccadic latency and inaccuracy. MK-0343 0.75 mg was equipotent with lorazepam as indicated by SPV (-42.4 deg/s), saccadic latency (0.02 s) and VAS alertness scores (1.50 ln mm), while effects on memory and postural stability were smaller. MK-0343 0.25 mg only affected postural stability to a similar extent as MK-0343 0.75 mg. The effect profile of MK-0343 0.75 mg is different from the full agonist lorazepam, which could reflect the selective actions of this compound. Although less effect on VAS alertness was expected, diminished effects on memory and postural stability were present. Clinical studies in anxiety patients should show whether this dose of MK-0343 is therapeutically effective with a different side-effect profile.