We investigated the effect of incision and sternotomy on the auditory evoked potential (AEP) and EEG, to try to predict a haemodynamic response to incision or sternotomy using the AEP and EEG in 41 patients undergoing cardiac surgery during propofol and alfentanil anaesthesia. The AEP and EEG were recorded before incision, between incision and sternotomy, and after sternotomy. Peak latencies and amplitudes of AEP peaks V, Na, Pa, Nb, Pb and Nc were determined. From the EEG the median, spectral edge and peak power frequencies, and percentages of delta, theta, alpha and beta power were calculated. Each patient was classified as responsive, equivocally responsive or unresponsive to incision or sternotomy based on increase in arterial pressure and heart rate on incision and sternotomy. Before incision, Nb and Pb latency and propofol concentration were higher for unresponsive patients but heart rate and median frequency before incision were lower. After sternotomy, Pa and Nb amplitude, peak power frequency and percentage alpha power were higher, and percentage theta power lower for responsive patients. Pa latency was higher after sternotomy for unresponsive patients. Using a combination of heart rate, arterial pressures and features derived from the AEP (all recorded before incision), the occurrence of a response to incision could be predicted in individual patients with a sensitivity of 85%, positive predictive accuracy of 63% and total accuracy of 72%. We conclude that AEP are more sensitive to pain stimuli than spectral features of the spontaneous EEG. In addition, the AEP may help in predicting inadequate anaesthesia.