The bispectral index, a value derived from the electroencephalogram, has been proposed as a measure of anaesthetic effect. The aim of the present study was to evaluate the bispectral index during midazolam–fentanyl anaesthesia for cardiac surgery for its possible role as a predictor of increases in systolic blood pressure during endotracheal intubation and sternotomy. After institutional approval 15 consenting patients, scheduled for elective cardiac surgery, were selected for the study. Anaesthesia was induced in all patients with a loading dose of fentanyl 7.5–10 μg kg−1, midazolam 0.15 mg kg−1 and pancuronium 0.1 mg kg−1. After a further bolus dose of fentanyl 10–12.5 μg kg−1 prior to the start of incision and sternotomy, maintenance infusion rates of fentanyl 4–6 μg kg−1 h−1 and midazolam 0.1 mg kg−1 h−1 were started and continued through surgery at the discretion of the anaesthetist and guided by the presenting clinical and haemodynamic responses. The control of anaesthesia was never based on the value of the bispectral index. The mean bispectral index value decreased from 95.7 (3.1) at base-line to 59.5 (12.0) after induction of anaesthesia and then remained below 70 throughout surgery. However, there was an important interindividual variability in bispectral index values despite standardized dosages of fentanyl and midazolam. There was no significant correlation between the bispectral index values in the pre-intubation and pre-incision period and the changes in systolic blood pressure during endotracheal intubation and sternotomy, respectively. In conclusion, the large intersubject variability in the bispectral index values should be investigated further in the light of the great variability in the clinical effects of midazolam and fentanyl. The lack of significant correlation between the bispectral index values and the haemodynamic responses suggest that the bispectral index, which is a helpful monitor of anaesthetic depth, is not a very reliable monitor of global anaesthetic adequacy during total intravenous anaesthesia with a combination of midazolam and fentanyl in cardiac surgical patients.