Combined somatosensory evoked potential (SEP) and regional brain oxygen saturation (rSO(2)) monitoring and simultaneous measurement of plasma levels of S100Beta and creatine kinase-isozyme BB (CK-BB) were performed to evaluate how reliable these diagnostic modality complexes are in the early prediction of neurological complications after surgery. Between 1999 and 2002, intraoperative SEP and rSO(2) monitoring combined with measurements of S100Beta and CK-BB levels in blood were performed in 82 consecutive patients undergoing cardiovascular operations with cardiopulmonary bypass (CPB). Twelve (14.6%) of these patients were diagnosed as having neurological complications after surgery; seven with transient neurological dysfunction (8.5%), and five with permanent stroke (6.1%). Twenty one of 82 patients in whom rSO(2) was recorded were judged abnormal; however, only nine of the 21 (42.9%) were diagnosed as having brain damage - diagnostic sensitivity and specificity being 75.0% and 82.9%, respectively. All six patients who showed abnormal SEP during surgery had neurological complications, but normal SEP was recorded in six other patients with apparent evidence of neurological complications - diagnostic sensitivity and specificity being 50% and 100%, respectively. There were no significant differences in S100Beta levels between patients with and without brain complications at 1 h and 24 h after CPB, but significant differences were detected in CK-BB levels at 24 h after CPB. In conclusion, simultaneous abnormalities detected in SEP and rSO(2) are highly predictive of cerebral neurocirculatory disturbances, but they are not so sensitive in diagnosing restricted focal cerebral lesions. Additional determinations of blood CK-BB levels might be valuable only to confirm the newly established brain complications.