Blood flow during cardiopulmonary bypass (CPB) is calculated on body surface area (BSA). Increasing comorbidity, age and weight of today's cardiac patients question this calculation as it may not reflect individual metabolic requirement. The hypothesis was that a measured cardiac index (CI) prior to normothermic CPB is a better estimate. A cross-over study, with random allocation to CPB blood flow for 20 minutes based on either a calculation (2.4 L/min/m(2)) or on CI, with a switch to the opposite flow for another 20 minutes, was performed. Twenty-two elective cardiac surgery patients with normal ventricular function were included. Effect parameters were cerebral oxygenation, mixed venous saturation and arterial lactate. CI varied from 1.9 to 3.1 L/min/m(2) (median 2.4 L/min/m(2)). No differences in effect parameters were seen. In conclusion, a CPB blood flow based on an individual estimate did not improve cerebral and systemic oxygenation compared to a blood flow based on BSA.