Favorable neurological survival in out-of-hospital cardiac arrest (OOHCA) may be influenced by cerebral perfusion during resuscitation. Cerebral oximetry (COx) provides a portable, noninvasive, real-time index of cerebral perfusion that has not been studied in OOHCA. This study examined the feasibility of using COx to measure cerebral perfusion during OOHCA. As a secondary aim, we tested the hypothesis that cerebral perfusion, measured by COx, would decrease with hyperventilation. Subjects were patients with medical OOHCA. A physician responded to the scene of cardiac arrest calls and applied an INVOS 3000 COx probe (Somanetics) to the frontal skull. In a cross-over design, readings were recorded for 2 min while ventilation rate was maintained at 10/min, then for 2 min at 24/min. COx readings were recorded by the oximeter and manually by the investigator. Statistical analyses were done using a paired t-test. Sixteen subjects were enrolled, four had return of pulses. COx reliably detected cortical oxygenation in only one subject during cardiac arrest. None of 16 patients exhibited consistently detectable levels of oxygen during cardiac arrest. In three subjects with ROSC, readings increased with return of pulses and with increasing blood pressure. In a fourth subject the protocol was completed and the device removed, subsequently ROSC was noted and the device re-applied. No patient exhibited any change in oxygen levels with variation of ventilation rates during CPR. The use of cerebral oximetry during OOHCA is feasible. In our sample of OOHCA patients, cerebral perfusion is rarely detectable using an oximeter during CPR. Ventilation rate does not alter the oximeter readings. It is possible that the current standard mechanical method of cardiopulmonary resuscitation provides little or no cerebral oxygenation during OOHCA.