The light- and electron microscopic localization of glutamate and glutamine in the rat organ of Corti was studied by means of antisera raised against the respective amino acids coupled to carrier proteins. The light microscopic analysis was performed in semithin sections treated according to the peroxidase-antiperoxidase procedure. The two amino acids were visualized in the same ultrathin sections by use of postembedding immunocytochemistry with two different gold particle sizes. The distribution of aspartate-like immunoreactivity was also recorded, but only at the light microscopic level. In the hair cells, the level of glutamate-like immunoreactivity was higher than that in supporting cells but lower than that in the presumed glutamatergic terminals of cerebellar parallel and mossy fibres. The latter types of terminal were sampled from ultrathin sections that had been incubated under the same conditions as the cochlear sections. Within the hair cells, gold particles signalling glutamate were enriched on mitochondria but not on clusters of synaptic vesicles. Glutamine-like immunoreactivity was present in hair cells as well as supporting cells. The glutamate/glutamine ratio, expressed as the ratio between the respective gold particle densities, was considerably lower for hair cells compared with the cerebellar excitatory terminals. No consistent difference was found between outer and inner hair cells in relation to the levels and subcellular distribution of glutamate and glutamine immunoreactivities. Aspartate-like immunoreactivity was accumulated in outer hair cells, with some labelling also of border cells and Böttcher cells. While the present study confirmed the presence of glutamate in hair cells and demonstrated that these cells are also endowed with the important glutamate precursor glutamine, it revealed notable differences between hair cells and presumed glutamatergic terminals in the CNS. These could reflect differences in the synthesis and compartmentation of transmitter glutamate. Methodological factors could also contribute. Alternatively, the differences could be interpreted to suggest that the hair cell transmitter is not glutamate, but a similar compound. Aspartate could be a candidate in the case of the outer hair cells.