Although glucocorticoids are sometimes used for the treatment of vertigo in certain disorders such as Meniere's disease, the mechanism underlying anti-vertigo effect remains unknown. The present study was performed to examine the effects of a glucocorticoid, dexamethasone, on neuronal activity in the medial vestibular nucleus (MVN) to determine whether or not the drug acts directly on the MVN neuron using alpha-chloralose-anesthetized cats which were fixed in a stereotaxic instrument placed on a turn-table. Single neuronal activities in the MVN were extracellularly recorded with a glass-insulated silver wire microelectrode attached along a seven-barreled micropipette. Each pipette was filled with dexamethasone phosphate (0.1 M), monosodium glutamate (1 M), glutamic acid diethylester (GDEE) (0.05 M: a non-selective glutamate receptor antagonist), CoCl2, (0.1 M: a non-specific calcium channel blocker), RU38486 (0.01 M: glucocorticoid receptor antagonist) or potassium canrenoate (0.1 M: a mineralo-cortical receptor antagonist). These chemicals were microiontophoretically applied to the immediate vicinity of the target neuron being recorded. The effects of the drugs were examined on type I neurons which were identified according to responses to rotation: the neuron showed an increase and a decrease in firing with ipsilateral and contralateral rotation to the recording site, respectively. Microiontophoretically applied dexamethasone (50-200 nA) dose-dependently increased spontaneous firing of MVN neurons. However iontophoretic application of GDEE did not affect the dexamethasone-induced increase in firing of the MVN neurons but inhibited glutamate- and rotation-induced firing. Microiontophoretically applied Co2+ did not affect dexamethasone-, glutamate- and rotation-induced firing. However, dexamethasone-induced firing was dose-dependently suppressed by iontophoretic RU38486, but not by canrenoate. Then a microdialysis study using alpha-chloralose-anesthetized cats was performed to determine whether or not dexamethasone affects the release of glutamate from vestibular nerve terminals. The microdialysis probe (CMA/10, 2 mm) was inserted into the MVN and perfused with Ringer solution at 2 ml/min. Samples were collected at 10-min intervals. Endogenous glutamate was measured using the HPLC-ECD method. When repetitive stimuli (200 microseconds duration, 0.5 mA and 5 Hz) were given to the vestibular nerve for 10 min, an increase in the release of glutamate was observed. Dexamethasone did not produce spontaneous or stimulation-induced release of glutamate. These results suggest that dexamethasone acts directly on the MVN neuron to excite neuronal activity through glucocorticoid receptors on neuron membranes, but the excitation is not due to the release of glutamate.