Objective To compare and evaluate the sedation effect and safety of dexmedetomidine and midazolam in lower limbs of orthopaedic surgery. Methods Fifty-two patients undergoing lower limbs surgery were randomly assigned to either group D (n=27, dexmedetomidine, 0.5μg/kg/hr, loading dose 0.5μg/kg) or group M (n=25, midazolam, 50μg/kg/hr, loading dose 30μg/kg). An infusion was administered after mounting a BIS monitor and combined spinal-epidural anesthesia. The target BIS level was ≤85. An additional bolus dose of the study drug or cessation of the infusion was adjusted according to the BIS level, and the change should be recorded. Respiratory rate (RR), heart rate (HR), pulse oxygen saturation (SpO2), mean arterial blood pressure (MAP) were continually monitored. Sedation should be evaluated according to the Ramsay sedation scale before the drug withdrawal. Result HR at 30 minutes after infusion in group D was significantly lower than that in group M, P <0.05. There was no statistic difference in the RR, MAP and RSS score between two groups. Dose adjustments were required in six and seven patients in Groups D and M, respectively. Five patients needed to raise the mandible because of the declining of SpO2 by glossocoma. Conclusion Dexmedetomidine infusion mildly decreased heart rate in the later periods of lower limbs surgery, while few affected to the MAP. Under the same sedation, dexmedetomidine compared to midazolam may have more advantages in maintaining the unobstructed of the respiratory tract, and may be more suitable for the sedation of this surgery.