Background Coma after cardiopulmonary resuscitation (CPR) is commonly seen in daily clinical practice. How to objectively evaluate brain function after CPR is essential to the following treatment. Coma patients after CPR had been studied prospectively at the Neuro-Intensive Care Unit of Xuanwu Hospital since 2002. In this study, we focused on the topic of how to evaluate the severity of coma after CPR. Methods From April 2002 to November 2004, patients in coma 24 hours after CPR were monitored, the evaluation methods included Glasgow coma score (GCS) brain stem reflection, and spinal reflection. Laboratory evaluation included electroencephalography (EEG), brainstem auditory evoked potential (BAEP), short latency somatosensory evoked potential (SLSEP), and transcranial Doppler (TCD). Results Twenty-four of 35 patients (68.57%) were in deep coma. The GCS was 3 except for 2 patients; EEG was evaluated not less than grade IV except for 4 patients, BAEP was evaluated as grade III except for 3 patients, and SLSEP was evaluated as grade III except for 1 patient. Twenty-four patients died within 1 month and 11 of them (45.83%) were determined as brain death. Glasgow outcome score (GOS) was evaluated as grade I. Eleven of the 35 patients survived and their consciousness changed from deep coma to coma vigil. EEG was evaluated as grade I in 5 patients, BAEP and SLSEP were evaluated as grade I in 3 patients, and GOS was all evaluated as grade II among the 11 patients. Two patients (18.18%) regained consciousness in 35 and 90 days after cardiopulmonary resuscitation and GOS was evaluated as grade IV and III, respectively. Conclusion Combined or continuous evaluation of clinical examinations and laboratory tests can accurately and objectively determine brain function after CPR.