We sought to investigate the effects of nitrous oxide on intraoperative electrocorticogram (ECoG) spike activities in 11 patients with intractable epilepsy. Grid electrodes were placed on the brain surface, and ECoG was recorded under the following conditions: 1.5 minimal alveolar anesthetic concentration (MAC) sevoflurane without nitrous oxide and 1.5 MAC sevoflurane with 50% nitrous oxide. The number of spikes for 5 minutes and the percentage of leads with spikes of total leads measured were assessed in each condition. The median numbers (25-75th) of spikes without and with nitrous oxide were 127 (87-368) and 61 (43-247), respectively. The numbers of spikes with nitrous oxide were significantly lower than those without nitrous oxide (P<0.05). The median percentages of leads with spikes without and with nitrous oxide were 68 (25-81) and 61 (28-70), respectively, and there were no significant differences in percentages of leads with spikes between the conditions. These results indicate that nitrous oxide attenuated the frequency of spikes on ECoG in epileptic patients, although it did not affect the extent of areas with spike activity.