Potassium accumulation or impaired potassium clearance has been hypothesized to contribute to epileptogenesis in gliotic epileptogenic foci. To test this hypothesis, potassium clearance rates following direct cortical stimulation were measured in the cortex of monkeys rendered epileptic by the injection of alumina gel into the motor area. Reactive gliosis at the sites in which potassium clearance was measured was then quantitated histologically and compared with potassium clearance rates. Dense gliosis was associated with slowed potassium clearance, although the base-line potassium level appeared no different in actively epileptogenic areas or gliotic areas compared with normal areas. Possible mechanisms and significance of slowed potassium clearance in the alumina focus are discussed.