Rat cerebellar Purkinje cells show a high level of taurine-like immunoreactivity. Light-microscopic immunocytochemistry indicated that the level of taurine in these cells was substantially decreased in animals that had survived for 4 h after an intraperitoneal injection of distilled water. This treatment resulted in a 15-20% reduction in plasma osmolality. The changes in the Purkinje cells were accompanied by an increased immunolabeling of neighboring glial cells (Golgi epithelial cells). The changes in both cell types were reversed in animals whose plasma osmolality had been normalized by injections of hypertonic saline 4 h after the water loading. Adjacent sections incubated with a GABA antiserum did not exhibit any overt changes in response to the hypo-osmotic stress. Quantitative electron-microscopic analysis of ultrathin sections subjected to postembedding immunogold cytochemistry indicated that the Purkinje cells had lost 50-60% of their taurine contents after water loading and that the loss affected all intracellular compartments, including mitochondria and cytoplasmic matrix. The loss of taurine immunoreactivity from Purkinje cells was accompanied by an estimated 70-80% increase in the contents of immunoreactive taurine in adjacent glial cells. Biochemical recordings of tissue amino acids in a parallel series of animals revealed a 12% reduction in cerebellar taurine contents 4 h after water loading (value corrected for changes in specific gravity). This reduction had progressed to 32% after 8 h and was only partly prevented by normalization of plasma osmolality. The tissue levels of GABA and several other amino acids showed a decrease similar to that of taurine, while glutamine displayed a considerable increase after water loading. Our findings indicate that acute reductions in plasma osmolality cause a flux of taurine from Purkinje cells to glia, and that this flux is reversed upon normalization of plasma osmolality. These changes are superimposed on a decrease in the biochemically recorded tissue level of taurine. Unlike the cellular redistribution, this decrease was not reversible within the time frame of the present study, and it was not specific for taurine. Cellular redistribution of taurine may represent a rapid adjustment to osmotic perturbations in vivo. In addition, it may reflect a higher priority for neuronal compared with glial volume regulation.