Objective: Reversal of brain blood flow is necessary for retrograde cerebral perfusion (RCP) to have any metabolic benefit, but RCP is commonly used with little clinical evidence of the true incidence of reverse brain blood flow and impact. S-100B is exclusive to the brain and spinal cord and released during hypothermic circulatory arrest (HCA). True reverse brain blood flow (tRBBF) during RCP may be determined by demonstrating an excess of S-100B in the effluent blood from the common carotid artery compared to blood entering the brain. Methods: Simultaneous blood samples were drawn from the jugular bulb and left common carotid artery during RCP at 5 min intervals in ten patients undergoing aortic surgery, utilizing HCA and subjected to blood gas, glucose and S-100B quantification. RCP was instituted at maximum pressure of 25 mmHg. Trans-cranial Doppler (TCD) continuously monitored the middle cerebral artery velocity (MCAV). Results: The mean HCA and RCP durations were 31 min (20-50 min). Reversal of MCAV was demonstrated in only 6/10 cases (mean, 6 cm/s). Veno-arterial (V-A) extraction of oxygen and glucose occurred in all cases (P<0.001), with the mean effluent pO 2 falling to 14 mmHg. V-A S-100 gradients greater than 5% were demonstrated in 8/10 cases and correlated with higher oxygen extraction (P<0.01). In patients with and without MCAV reversal, the S-100 gradients were 1.7 and 0.3 μmol/l, respectively (P<0.01). Conclusions: tRBBF occurred in nearly all patients. MCAV reversal appears to be a specific but insensitive guide to reverse perfusion. The de-saturation of effluent blood is not a reliable guide to true brain perfusion, and despite RCP, the brain remains ischaemic.