Objective To compare the brain protection by the combination of dexmedetomidine with ketamine or remifentanil in the patients after cardiac valve replacement(CVR) under cardiopulmonary bypass. Methods Ninety patients were selected for the study, who were transferred to ICU for sedation after elective CVR under cardiopulmonary bypass in the Anesthesiology Department of Xuzhou Medical University Hospital from March 2017 to February 2018. The patients were divided into dexmedetomidine-ketamine group (DK group) and dexmedetomidine-remifentanil group (DR group), with 45 patients in each group. Both groups received dexmedetomidine 0.2-0.7 μg/(kg·h) via intravenous pumping, but ketamine 0.3-0.5 mg/(kg·h) was administered in the DK group and remifentanil 2.5-5.0 μg/(kg·h) in DR group via intravenous pumping. The patients were monitored for the hemodynamic parametersincluding the mean arterial pressure (MAP) and mean heart rate (MHR). The brain injury severity indices including central nerve specific protein (S100β) and neuron-specific enolase (NSE) were measured before sedation (T₀), and 2 h (T₁), 6 h (T₂), 12 h (T₃) and 24 h (T₄) after sedation. Records were kept of the delirium incidence, dosage of dexmedetomidine, haloperidol and dezocine, awakening time, extubation time, duration of ICU stay, postoperative hospital stay and incidence of adverse events. SPSS statistics 19.0 was used for data processing. Student′s t test, Chi-square test or rank-sum test were used for intergroup comparison, and repeated measurement variance analysis for intragroup.Results The levels of S100β and NSE at T1-4 in both groups were significantly lower than those at T₀ (P<0.05). The decrease of S100β at T1-4 and the NSE at T1-3 in the DK group was greater than that in the DR group, difference being statically significant (P<0.05). There was no significant difference in MAP and MHR between 2 groups before sedation. MAP at 30 min-9 h and MHR 30 min-8 h after sedation in the DK group were significantly higher than those in the DR group, and the differences were statistically significant (P<0.01). The fluctuation of MAP and MHR in the DK group was more moderate than in the group DR. During sedation, the recordings in the DK group were significantly lower than those in the DR group of the concentration of dexmedetomidine [(374.3±52.7) vs (504.6±69.3)μg], dezocine [(9.6±2.7) vs (15.6±3.3)mg] and haloperidol [(29.7±3.2) vs (46.6±3.4)mg]; occurrences of hypotension [6.6%(3/45) vs 24.4%(11/45)] and bradycardia [4.4%(2/45) vs 20.0%(9/45)]; and awakening time [(457.2±14.5) vs (504.2±16.9)min] and extubation time [(511.7±15.9) vs (553.5±20.9)min] (P<0.05), but no significant differences were observed in delirium ratio, nausea/vomiting, ICU stay time and postoperative hospital stay (P＞0.05). Conclusion Dexmedetomidin-ketamine and dexmedetomidin-remifentanil have similar sedative and analgesic effects after CVR with cardiopulmonary bypass, but the former provides more significant cerebral protection with more reasonable and stable hemodynamics and significantly decreased occurrences of adverse reactions.