Background Brain death (BD), the irreversible cessation of function in the whole brain, is a well-known condition in most countries. The criteria and practical guidelines for brain death determination (BDD) in China were issued by the Brain Injury Evaluation Quality Control Center (BQCC) of the National Health and Family Planning Commission in 2013. Thereafter, we proposed a plan called the three-step quality control plan (three-step QCP) to ensure the safety and consistency of the clinical judgments regarding BD. By retrospectively reviewing this plan, we aimed to identify problems during its implementation and to provide suggestions for future work on quality control for BDD. Methods Data were retrieved from the BQCC database. The characteristics and test results of physicians undergoing a BDD training course and the BD case records submitted by hospitals before and after receiving accreditation were analyzed. Results In the first step of the plan, the error rate for physicians undergoing the BDD paper test was highest for limb movement discrimination (26.29%); this error rate was correlated with age (per 10-year increase) (odds ratio = 1.262, 95% confidence interval 1.067-1.491, P = 0.007) but was nonsignificantly associated with sex, specialty category, professional level, and hospital level (P > 0.05). During the second step of the plan, the highest percentage of problems was associated with apnea testing (22.75%), followed by ancillary testing of BDD (16.17%). In the last step, the highest percentage of problems in the case records was associated with apnea testing (41.73%). Conclusions The three-step QCP is of significant utility for ensuring accuracy and appropriateness in BDD. Simultaneously, this study provides important evidence for advancing quality control for BDD in the next stage.