Little information is available on the specific roles of different cellular mechanisms involved in extracellular K(+) homeostasis during neuronal activity in situ. These studies have been hampered by the lack of an adequate experimental paradigm able to separate K(+)-buffering activity from the superimposed extrusion of K(+) from variably active neurons. We have devised a new protocol that allows for such an analysis. We used paired field- and K(+)-selective microelectrode recordings from CA3 stratum pyramidale during maximal Schaffer collateral stimulation in the presence of excitatory synapse blockade to evoke purely antidromic spikes in CA3. Under these conditions of controlled neuronal firing, we studied the [K(+)]o baseline during 0.05 Hz stimulation, and the accumulation and rate of recovery of extracellular K(+) at higher frequency stimulation (1-3 Hz). In the first set of experiments, we showed that neuronal hyperpolarization by extracellular application of ZD7288 (11 microM), a selective blocker of neuronal I(h) currents, does not affect the dynamics of extracellular K(+). This indicates that the K(+) dynamics evoked by controlled pyramidal cell firing do not depend on neuronal membrane potential, but only on the balance between K(+) extruded by firing neurons and K(+) buffered by neuronal and glial mechanisms. In the second set of experiments, we showed that di-hydro-ouabain (5 microM), a selective blocker of the Na(+)/K(+)-pump, yields an elevation of baseline [K(+)]o and abolishes the K(+) recovery during higher frequency stimulation and its undershoot during the ensuing period. In the third set of experiments, we showed that Ba(2+) (200 microM), a selective blocker of inwardly rectifying K(+) channels (KIR), does not affect the posttetanus rate of recovery of [K(+)]o, nor does it affect the rate of K(+) recovery during high-frequency stimulation. It does, however, cause an elevation of baseline [K(+)]o and an increase in the amplitude of the ensuing undershoot. We show for the first time that it is possible to differentiate the specific roles of Na(+)/K(+)-pump and KIR channels in buffering extracellular K(+). Neuronal and glial Na(+)/K(+)-pumps are involved in setting baseline [K(+)]o levels, determining the rate of its recovery during sustained high-frequency firing, and determining its postactivity undershoot. Conversely, glial KIR channels are involved in the regulation of baseline levels of K(+), and in decreasing the amplitude of the postactivity [K(+)]o undershoot, but do not affect the rate of K(+) clearance during neuronal firing. The results presented provide new insights into the specific physiological role of glial KIR channels in extracellular K(+) homeostasis.