Prolongation of peripheral nerve blockade by intravenous dexamethasone may be extended by intravenous dexmedetomidine. We randomly allocated 122 participants who had intravenous dexamethasone 0.15 mg.kg ^-1 before interscalene brachial plexus block for day-case arthroscopic rotator cuff repair to intravenous saline (62 participants) or intravenous dexmedetomidine 1 μg.kg ^-1 (60 participants). The primary outcome was time from block to first oral morphine intake during the first 48 postoperative hours. Fifty-nine participants reported taking oral morphine, 25/62 after placebo and 34/60 after dexmedetomidine, p = 0.10. The time to morphine intake was shorter after dexmedetomidine, hazard ratio (95%CI) 1.68 (1.00-2.82), p = 0.049. Median (IQR [range]) morphine doses were 0 (0-12.5 [0-50]) mg after control vs. 10 (0-30 [0-50]) after dexmedetomidine, a difference (95%CI) of 7 (0-10) mg, p = 0.056. There was no effect of dexmedetomidine on pain at rest or on movement. Intra-operative hypotension was recorded for 27/62 and 50/60 participants after placebo vs. dexmedetomidine, respectively, p < 0.001. Other outcomes were similar, including durations of sensory and motor block. In conclusion, dexmedetomidine shortened the time to oral morphine consumption after interscalene block combined with dexamethasone and caused intra-operative hypotension.