Extracellular potassium concentrations, [K+]e, were measured in vivo in the rat dorsal hippocampus using valinomycin-based double-barrelled ion-selective microelectrodes. Experiments were conducted under chloral hydrate anaesthesia. The microelectrodes were implanted stereotaxically, after which different gas mixtures were administered by inhalation. Transient hypoxia was induced by changing the inspired gas from 20% O2/80% N2 to 10-0% O2/90-100% N2 for 0.5-2 min. Resting [K+]e in the dorsal hippocampus was 3.4 +/- 0.09 mM; 0.5, 1 or 2 min of 100% N2 administration caused a rapid rise of [K+]e to 0.75, 1.9 and 15 mM, respectively. Following 0.5 min of 100% N2, the switch back to 20% O2/80% N2 produced an almost instantaneous return to normal levels. The return of [K+]e to basal levels was more delayed after 1 or 2 min of 100% N2 inhalation. The rise of hippocampal [K+]e induced by hypoxia was influenced by body temperature, the increase being five-fold higher in rats whose body temperature was raised from 33 to 37 degrees C using a heating blanket. Three potassium-channel blocking agents, quinine, 4-aminopyridine and gliquidone, were tested for their action on the increase in [K+]e, induced by inhalation of 100% N2 for 0.5 min. Both 4-aminopyridine and quinine, administered systemically, attenuated the anoxia-induced rise in [K+]e by 70 and 35%, respectively. In contrast, gliquidone, given by intracerebroventricular injection, had no effect, suggesting that ATP-sensitive potassium channels are not involved in this very early change in [K+]e.