We assessed the role of extracellular potassium ([K+]e) on the increase in cerebral blood flow (CBF) during hypoxia, and we tested whether it was affected by glibenclamide or ouabain. Cortical CBF was measured using the hydrogen clearance technique in enflurane-anesthetized rats, and local [K+]e was measured with K+ microelectrodes adjacent to the hydrogen electrode. Eucapnic hypoxia (arterial Po2 approximately 35-40 Torr) increased CBF twofold and caused a modest rise in [K+]e (from 2.9 +/- 0.2 to 3.7 +/- 0.2 mM; mean arterial blood pressure, ABP, 86 +/- 5 mmHg). If ABP fell < 70 mmHg during hypoxia, no increase in CBF was seen, whereas [K+]e increased to > 20 mM. Glibenclamide (10-100 microM intracortically) attenuated [K+]e and CBF during hypoxia (ABP approximately 75 mmHg, P < 0.01). Ouabain (20-1,000 microM) increased [K+]e; however, it did not remove the hypoxic-induced rise in [K+]e. We conclude that glibenclamide-sensitive potassium channels contribute to the accumulation of [K+]e during hypoxia, although an increase in CBF during hypoxia can occur without a marked rise in [K+]e. Furthermore, if ABP falls below the lower limit of autoregulation during hypoxia, there is no increase in CBF, yet there is a large increase in [K+]e.