OBJECTIVE: The study's aim was to compare response entropy (RE) and state entropy (SE) with bispectral index (BIS) electroencephalography (EEG) as an alternative cerebral monitoring tool in patients scheduled for coronary artery bypass graft surgery. DESIGN: Prospective, observational single-center study. SETTING: University hospital. PARTICIPANTS: Thirty patients undergoing coronary artery bypass graft surgery receiving remifentanil-propofol anesthesia. INTERVENTIONS: Surgery was performed with cardiopulmonary bypass (CPB) and cardiac arrest in 15 patients, with CPB without cardiac arrest in 9 patients and without CPB in 6 patients. MEASUREMENTS AND MAIN RESULTS: RE, SE, BIS, burst suppression ratio (BSR), and frontal electromyography (f-EMG) were detected simultaneously. RE and SE compared favorably with BIS and their correlations were strong (r(2) = 0.6, r(2) = 0.55, respectively). The mean bias of RE and BIS was -1.8, but limits of agreement were high (+20.5/-24.1). RE and SE tended to be lower than the BIS values in the CPB subgroups. The detection of BSR was similar with RE and SE and the BIS. A strong correlation existed between BIS and f-EMG (r(2) = 0.62) in contrast to RE (r(2) = 0.45) and SE (r(2) =0.39). BIS monitoring was significantly more disturbed than RE and SE with 9.1% +/-10.9% and 0.1% +/- 0.2% of the total anesthesia time, respectively. Neither implicit nor explicit memory was shown. CONCLUSION: RE and SE are comparable with the BIS but showed significantly less interference from f-EMG and superior resistance against artifacts. Thus, spectral entropy is more suitable than the BIS during propofol-remifentanil anesthesia in cardiac surgery patients.