Agonists at the benzodiazepine receptor (BZR) produce their effects through potentiation of the inhibitory alpha-aminobutyric acid-mediated neurotransmission in the central nervous system via positive allosteric modulation of the gamma-aminobutyric acidA receptor. Agonists with high intrinsic efficacy are anticonvulsant, anxiolytic, muscle relaxant and sedative, whereas agonists with low intrinsic efficacy (partial agonists) are predominantly anticonvulsant and anxiolytic, but antagonize muscle relaxant and sedative effects of full agonists. The four BZR ligands triazolam, diazepam, Ro 19-8022 (a benzoquinolizinone) and bretazenil (Ro 16-6028, an imidazobenzodiazepinone) were pharmacologically characterized in various neurological and behavioral paradigms in mice: two anticonvulsant tests (prevention of audiogenic and pentylenetetrazol-induced seizures), a conflict test which reveals both anxiolytic and sedative properties and two tests which mainly measure motor impairment (rotarod and horizontal wire test). Although triazolam and diazepam elicited an effect in all tests, Ro 19-8022 and bretazenil exhibited anticonvulsant and anxiolytic properties, but virtually failed to induce motor impairment and severe sedation. In separate experiments, fractional BZR occupancy in vivo was assessed by inhibition of [3H]flumazenil binding and correlated with the pharmacological effects. Although diazepam and triazolam produced effects beginning at low to intermediate fractional BZR occupancy, Ro 19-8022 and bretazenil required a higher BZR occupancy to do so, in accordance with their partial agonistic character. With the two full agonists, anticonvulsant and anticonflict activities were elicited at a lower fractional BZR occupancy than muscle relaxant and sedative effects.(ABSTRACT TRUNCATED AT 250 WORDS)