Currently the most used perfusion techniques during aortic arch surgery to prevent cerebral damage include hypothermic circulatory arrest, retrograde cerebral perfusion and selective cerebral perfusion (SCP). The application of simplified SCP, which does not require deep hypothermia, has become an alternative procedure for brain protection. Including the physiological principle of autoregulated cerebral blood flow, cerebral perfusion flow is not predetermined, but differentiated from the different cannula sizes for the lower and upper body perfusion. In a mock circulation loop, we could show that resistance changes in the two compartments led to flow shifts between the systemic and brachiocephalic regions. In addition to mechanical factors cerebral perfusion is determined from physiological changes. In practice, these shifts can be initiated with disrupted autoregulation due to ph-stat management or dramatic pressure changes. To prevent mismatched cerebral perfusion extended perioperative monitoring was included in our clinical setting. With bilateral somatosensory evoked potentials, a computer aided topographical electroencephalometry system, transcranial doppler-sonography and jugular venous bulb saturation, we could provide a sufficient bihemispheric perfusion. Between 1990 and 1995 we operated on 21 patients using SCP. Intraoperatively no signs of cerebral ischaemia due to inadequate perfusion could be observed. Only temporary neurological changes were found postoperatively. In summary, the simplified SCP, despite its physiological basis, is intricately involved in control and influence. We think that the application of SCP is safe if extended neurophysiological monitoring is included in the clinical setting.