Deep hypothermia is an effective technique for neuroprotection in cardiac surgery. However, standard body temperature measurement may deviate from actual brain temperature. Therefore, we simultaneously measured brain and core temperatures during neurosurgical interventions in hypothermic circulatory arrest to determine its accuracy. Between 1994 and May 2007, 26 patients (12 female, mean age 46+/-14 years), with complex intracranial aneurysms, underwent resection or clipping applying closed chest cardiopulmonary bypass and hypothermic circulatory arrest via inguinal cannulation. During surgery, temperature probes were positioned in the brain, tympanum, bladder, rectum and pulmonary artery. Mean cardiopulmonary bypass time was 147+/-39 min, mean circulatory arrest time was 28+/-8 min. Brain temperatures were best reflected by bladder and tympanum probes (Pearson's correlation coefficients: bladder=0.83; tympanum=0.80; pulmonary artery=0.63; rectum=0.37; P<0.05). Mean deviations from brain temperature were +0.2+/-2.7 degrees C at the tympanum, -0.8+/-2.6 degrees C in the bladder, -0.7+/-2.6 degrees C in the pulmonary artery and -1.8+/-4.4 degrees C in the rectum. In conclusion, temperature monitoring in the bladder and tympanum reliably reflects brain temperature. Temperature measurements in the pulmonary artery and rectum are less optimal.