The sensitivity and regional specificity of intraoperative electroencephalographic (EEG) monitoring in cerebral ischemia was evaluated in a new experimental model of temporary focal cerebral ischemia in rabbits. EEG potentials were recorded directly from the cortical surface using a bipolar disc electrode grid and were analyzed by computer. Groups of 5 animals each underwent temporary occlusion of the left middle cerebral arterial trunk for 5, 10, 20, 30, 45, or 60 minutes. EEG data were recorded from the cortex proximal (temporal site) and distal (parasagittal site) to the middle cerebral arterial trunk during occlusion and 2 hours of reperfusion. EEG suppression was detected immediately after occlusion at the temporal site by analysis of power spectra in 29 of 30 rabbits (mean power, 32% of base line), by compressed spectral array (CSA) edge analysis in 23, and by analysis of the conventional EEG wave form in 24. Within 5 minutes after the start of occlusion, all 30 rabbits showed EEG power suppression and 26 showed decrease in the CSA edge frequency or in the routine EEG wave form. By the end of the occlusion period, EEG power at the temporal site had decreased to 20.5% of base line. At the parasagittal site, a lesser degree of EEG suppression was detected; 20 rabbits had an initial loss of EEG power (mean, 85.7% of base line), 13 had decrease in the CSA edge, and 7 had suppression of the EEG wave form. By the end of the occlusion period, spectral power at the parasagittal site had decreased in 25 of 30 rabbits to a mean of 86.9% of base line.(ABSTRACT TRUNCATED AT 250 WORDS)