Abstract: Objective To investigate the feasibility of mean arterial pressure (MAP) as feedback variable of propofol closed-loop target controlled infusion in neurosurgical anesthesia. Methods 46 patients for elective brain tumor operation were divided into target controlled group and closed-loop group randomly,23 patients in each group. The MAP feedback value was set at 70 mm Hg in closed-loop group. The plasma target two groups. The value of MAP, HR, Heart rate variation index (HRVI), Bispectral index of EEG (BIS), and propofol dosage were recorded at the time of before anesthesia,intubation,5 min after intubation, skin incision,skull open,pallium procedure, 1 h tumor procedure, and skin suture. Results The descensus of MAP in targetcontrolled group was significant than closed-loop group at the time of 5 min after intubation, pallium procedure,and 1 h tumor procedure(P<0.05). The descensus of HR in two groups were similar,and no statistic difference (P > 0.05), HRVI was the same. The BIS value was in lower degree in most patients, and the value of target controlled group was significantly lower than that of close-loop group during the intracranial procedure period (P<0.05). Propofol total dosage (825.9±270.4) mg of closed-loop group was significantly less than target controlled group (1302.6±321.5) mg(P<0.01). Conclusion MAP was a safe feedback variance for adjusting propofol closed-loop target controlled infusion in craniocerebral operations,it is beneficial for stable circulation and less propofol total dosage.