Objective: To determine whether standard P50 after cardiac surgery decreases and whether decreased P50 is related to the transfusion of red blood cells (RBCs), acid-base changes, body temperature, oxygen parameters and/or duration of cardiopulmonary bypass (CPB). Design: Pilot study in cardiac surgery patients. Setting: University hospital. Patients: 12 consecutive elective cardiac surgery patients. Interventions: Blood was taken before surgery, after CPB and in the intensive care unit until 18 h postoperatively. Cardiac output and oxygen consumption were measured. Buffy coat-poor RBCs were transfused, anticoagulated with citrate-phosphate-dextrose buffer and stored in saline-adenine-glucose-mannitol at 4°C, when haemoglobin was < 5.6 mmol·l-1. Measurements and results: Standard P50 was calculated from measured partial pressure of oxygen and of carbon dioxide, pH and oxygen saturation in mixed venous blood (SvO2) using the Severinghaus formula. Median length of RBC storage was 25 days. Standard P50 after surgery was significantly lower than baseline value (p = 0.0001). The number of RBC units transfused and duration of CPB were conjointly associated with P50 (R2 = 0.72). Patients who received more RBCs consumed more oxygen. Conclusion: Cardiac surgery patients receiving more RBC units have lower standard P50 and consume more oxygen. P50 decreased more when the CPB took longer. Because a decrease in P50 implies a low ratio of mixed venous oxygen tension (PvO2) to SvO2, a shift in P50 should be taken into account when using SvO2 as a measure of global oxygen availability. When a direct measurement of SvO2 is not available, PvO2 should be used instead of calculated SvO2.