Ｏｂｊｅｃｔｉｖｅ　To compare the sedative effects of midazolam and different doses of dexmedetomidine(Dex) in patients undergoing branchial plexus block, and find out the suitable loading dose of Dex.　Methods　Eighty patients scheduled for upper limb operation were randomly divided into four groups(n=20): groups Dex(group D1, group D2, group D3) and group midazolam(group M). The patients in group D1, group D2, group D3 and group M were administered respectively with Dex 0.3, 0.5, 0.8 μg/kg and midazolam 0.05 mg/kg infusion 10 min before the branchial plexus block under the ultrasound guidance, followed by a maintenance infusion of Dex 0.5 μg·kg-1·h-1 or midazolam 0.05 mg·kg-1·h-1 respectively. The incidence of adverse effects and, the changes of MAP, HR, SpO2 and the scores of OAA/S before sedative administration(T0), 5 min after sedative administration(T1), just before the branchial plexus puncture(T2), just after the branchial plexus puncture(T3), the beginning of surgery(T4) and 30 min(T5), 40 min(T6), 60 min(T7) after sedative administration were recorded.　Results　At T5, SpO2 in group M were lower as compared to group Dex(P<0.05). Compared with group M, the scores of OAA/S were significantly higher in group D1 at T2-T7, while they were significantly lower in group D3 at T3-T5(P<0.05), but there were no significant differences in group D2(P>0.05). On the other side, the incidence of hypotension exhibited a positive correlation with the dose of Dex.　Conclusions　Dex can provide a better sedation than midazolam in this study which means intravenous injection of Dex in 0.5 μg/kg, followed by a maintenance infusion of 0.5 μg·kg-1·h-1 can offer effective sedation for the patients under branchial plexus block and stable hemodynamics as well as less adverse effects.