Background The brain death confirmation tests occupy a different position in each country's diagnostic criteria (or guideline); the choices of tests are also different. China brain death criteria include clinical judgment and confirmation tests. This study aimed to confirm the preferred confirmatory test and complementary confirmatory tests. Methods We did a clinical brain death determination on deep coma patients, and then divided them into brain death group and non-brain death group. According to the Chinese standards for determining brain death, both the groups accepted confirmatory tests including electroencephalograph (EEG), somatosensory evoked potentials (SEP), and transcranial Doppler (TCD). The sensitivity, specificity, false positive rate, and false negative rate were calculated to evaluate the accuracy of the confirmatory tests. Results Among the 131 cases of patients, 103 patients met the clinical criteria of brain death. Respiratory arrest provocation test was performed on 44 cases and 32 cases (73%) successfully completed and confirmed that they have no spontaneous breathing. Of the three confirmation tests, EEG had the highest completion rate (98%) and good sensitivity (83%) and specificity (97%); TCD had followed completion rate (54%) and not good sensitivity (73%) and specificity (75%); SEP had the lowest completion rate (49%), good sensitivity (100%), and not good specificity (78%). After the combination of SEP or TCD with EEG, the specificity can increase to 100%. Conclusions The completion rate of respiratory arrest provocation test remains a problem in the clinical diagnosis of brain death. If the test cannot be completed, whether to increase a confirmatory test is debatable. SEP had an ideal sensitivity, and the specificity will reach 100% after combining with TCD or EEG. When a confirmed test was uncertain, we suggest increasing another confirmatory test.