Patients with refractory seizures may undergo awake craniotomy and cortical resection of the seizure area, using intraoperative functional mapping and electrocorticography (ECoG). We used dexmedetomidine in 6 patients, transitioning successively from the asleep-awake-asleep method, through a combined propofol/dexmedetomidine sedative infusion, to dexmedetomidine as the only sedation. Initial experience with the asleep-awake-asleep method in 2 patients was successful with the replacement of propofol/laryngeal mask anesthesia, 20 to 30 minutes before ECoG testing, by dexmedetomidine infusion, maintained at 0.2 mcg kg-1 h-1 throughout neurocognitive testing. Propofol anesthesia was reintroduced for resection. One patient received combined dexmedetomidine (0.2 mcg kg-1 h-1) and propofol (200 mcg kg-1 min-1) infusions for sedation. Both infusions were stopped 15 minutes before ECoG. Subsequently, they were restarted and the epileptic foci resected. Three patients received dexmedetomidine as the sole sedative agent, together with scalp block local anesthesia, and incremental boluses totaling 150 to 175 mcg of fentanyl per case. Dexmedetomidine was started with 0.3 mcg kg-1 boluses and maintained with 0.2 to 0.7 mcg kg-1 h-1 for craniotomy, testing, and resection. The infusion was paused for 20 minutes in 1 patient to allow improvement in neurocognitive testing. This occurred within 10 minutes. All patients enjoyed good hemodynamic control, with blood pressure maintained within 20% of initial values, and made uneventful recoveries. The surgical conditions were all reported as favorable. Dexmedetomidine can be used singly for sedation in awake craniotomy requiring ECoG. Individual dose ranges vary, but a bolus of 0.3 mcg kg-1 with an infusion of 0.2 mcg kg-1 min-1 is a good starting point, allowing accurate mapping of epileptic foci and subsequent resection.