Neuronal excitation increases extracellular K(+) concentration ([K(+)]o) in vivo and in incubated brain tissue by stimulation of postsynaptic glutamatergic receptors and by channel-mediated K(+) release during action potentials. Convincing evidence exists that subsequent cellular K(+) reuptake occurs by active transport, normally mediated by Na(+),K(+)-ATPase. This enzyme is expressed both in neurons and in astrocytes but is stimulated by elevated [K(+)]o only in astrocytes. This might lead to an initial K(+) uptake in astrocytes, followed by Kir4.1-mediated release and neuronal reuptake. In cell culture experiments, K(+)-stimulated glycogenolysis is essential for operation of the astrocytic Na(+),K(+)-ATPase resulting from the requirement for glycogenolysis in a pathway leading to uptake of Na(+) for costimulation of its intracellular sodium-binding site. The astrocytic but not the neuronal Na(+),K(+)-ATPase is additionally stimulated by isoproterenol, a β-adrenergic agonist, but only at nonelevated [K(+)]o. This effect is also glycogenolysis dependent and might play a role during poststimulatory undershoots. Attempts to replicate dependence on glycogenolysis for K(+) reuptake in glutamate-stimulated brain slices showed similar [K(+)]o recovery half-lives in the absence and presence of the glycogenolysis inhibitor 1,4-dideoxy-1,4-imino-D-arabinitol. The undershoot was decreased, but to the same extent as an unexpected reduction of peak [K(+)]o increase. A potential explanation for this difference from the cell culture experiments is that astrocytic glutamate uptake might supply the cells with sufficient Na(+). Inhibition of action potential generation by tetrodotoxin caused only a marginal, nonsignificant decrease in stimulated [K(+)]o in brain slices, hindering the evaluation if K(+) reaccumulation after action potential propagation requires glycogenolysis in this preparation.