Adequate analgesia and sedation with adequate respiratory and hemodynamic control are needed during brain surgery in awake patients. In this study, a protocol using clonidine premedication, intraoperative propofol, remifentanil, and labetalol was evaluated prospectively in 25 patients (aged 50 +/- 16). In all but one patient, no significant problems regarding cooperation, brain swelling, or loss of control were noticed, and it was not necessary to prematurely discontinue any of the procedures. One patient, who was uncooperative and hypertensive, became apneic with increasing sedation, and needed a laryngeal mask airway inserted. Patients were hemodynamically stable; elevated systolic blood pressure (>or= 150 mm Hg) was measured infrequently, and there were no events of significant hypotension, tachycardia, or bradycardia. Events of hypoxemia (SAO2 <or= 95%), severe hypoxemia (SaO2 <or= 90%), or hypoventilation (respiratory rate <or=8 minute), were frequent in the first ten patients, but the incidence decreased significantly in subsequent patients (P < .001). Three patients developed a focal neurologic deficit, and two patients experienced intraoperative seizures. Nausea and vomiting were not recorded in any of the patients. Although these findings attest to the safety of awake craniotomy, they demonstrate the difficulty of achieving adequate sedation without compromising ventilation and oxygenation. The learning curve of using a new protocol and a new potent anesthetic drug is emphasized.