The relationship of bypass shunt blood flow to arterial pressure and stump pressure was studied in 12 informed patients undergoing carotid endarterectomy. These patients were anesthetized with halothane plus nitrous oxide. Systemic arterial blood pressure was measured with a radial artery catheter, and stump pressure as well as distal shunt pressure (carotid pressure distal to the shunt) were measured with a 25-gauge needle inserted cephalad to the bypass shunt. From the pressure drop across the bypass shunt, blood viscosity, geometry of the bypass tubing and the use of Poiseullie's law, the shunt blood flow can be calculated. The calculated shunt flow shows a close correlation with the difference between arterial pressure and stump pressure (r = 0.91) as well as with the slope of arterial pressure increase following carotid occlusion (r = 0.82); but it is poorly correlated with the stump pressure. Administration of 100 ml low molecular weight dextran solution does not improve the shunt blood flow, whereas 500 ml low molecular weight dextran significantly decreases blood viscosity. The patency of the carotid artery as assessed by angiogram does not give a proper indication of the need for a bypass shunt. The shunt flow as well as the need of such a shunt might be predicted by the use of the difference of arterial pressure and stump pressure. If the slope of arterial pressure increases rapidly following carotid occlusion, it can also be used to determine the necessity of such a shunt.