5-ethoxymethyl-7-fluoro-3-oxo-1,2,3,5-tetrahydrobenzo[4,5] imidazo[1,2a]pyridine-4-N-(2-fluorophenyl)carboxamide) (RWJ-51204) binds selectively and with high affinity (K(i) = 0.2-2 nM) to the benzodiazepine site on GABA(A) receptors. Considering the GABA shift, the intrinsic modulatory activity of RWJ-51204 is lower than that of full agonist anxiolytics (lorazepam, diazepam, alprazolam, and clonazepam) but similar to partial agonists (bretazenil, abecarnil, panadiplon, and imidazenil). RWJ-51204 was orally active in anxiolytic efficacy tests; pentylenetetrazole induced seizure inhibition in mice (ED(50) = 0.04 mg/kg), Vogel conflict in rats (ED(50) = 0.36 mg/kg), elevated plus-maze in rats (minimal effective dose = 0.1 mg/kg), and conflict in squirrel monkeys (ED(50) = 0.49 mg/kg). RWJ-51204 attenuated chlordiazepoxide-induced motor impairment in mice. Usually, RWJ-51204 was more potent than reference anxiolytics in rodent efficacy tests but less potent in monkey conflict. Usually, the slope of the dose-response lines for RWJ-51204 was more shallow than the full agonist anxiolytics but steeper than partial agonists in efficacy tests but typically shallow in tests for central nervous system side effects. In monkeys only mild or moderate sedation was observed at doses equivalent to 20 or 40 times the anxiolytic ED(50). RWJ-51204 fits into the partial agonist class of GABA(A) receptor modulators. In conclusion, RWJ-51204 exhibits a profile in in vitro experiments and in animal models, in mice and monkeys (but not in rats), suggesting that it has a profile of anxiolytic activity associated with less sedation, motor impairment, or muscle relaxation than currently available GABA(A) receptor modulators, i.e., the benzodiazepines.