Background We recently described noninvasive techniques to perform long-term DC-coupled EEG recordings at the bedside and have used these methods to demonstrate that inspection of infraslow ( Study Design and Methods Recordings were performed on patients receiving invasive monitoring for medically intractable localization-related epilepsy. Recordings were made by digital acquisition software using a commercial DC amplifier coupled to surgically placed subdural arrays of platinum electrodes, epilepsy with a total of 38 seizures in 9 patients captured during monitoring. The frequency bandwidth of record was limited from 0 to 70 Hz. The time of onset, location, amplitude, duration, and shape of infraslow signals up to 10 minutes prior to and 10 minutes after clinical onset of seizure (or electrical onset of seizure for subclinical seizures) were catalogued by a nonepileptologist (W.K.). Results Unequivocal infraslow signals, the majority having amplitude ranging from 800 to 4,000 μV in magnitude, were seen in 20 seizures in eight patients following the electrical ictal onset determined from the conventional recording (with a frequency band of 0.5 to 100 Hz), with a delay ranging from 5 seconds to 493 seconds. Of the 20 seizures with infraslow ictal activity, 16 localized to the same channel(s) as the ictal onset in conventional recording and one signal localized in a channel adjacent to the location of ictal onset. Conclusions Seizure localization using infraslow activity was concordant with that obtained from analysis of higher-frequency activity. The high voltage of this activity helps explain why infraslow activity can sometimes localize seizures better than conventional EEG on scalp recordings. We believe that the millivolt range shift observed in the DC-coupled EEG recordings is due to recruitment of sufficient volume of epileptogenic brain to elicit local changes in blood-brain barrier permeability or blood volume.