Intra-operative remifentanil is associated with increased postoperative analgesic requirements and opioid consumption. Dexmedetomidine has characteristics suggesting it may substitute for intra-operative remifentanil during general anaesthesia, but existing literature has reported conflicting results. We undertook this meta-analysis to investigate whether general anaesthesia including dexmedetomidine would result in less postoperative pain than general anaesthesia including remifentanil. The MEDLINE and PubMed electronic databases were searched up to October 2018. Only randomised trials including patients receiving general anaesthesia and comparing dexmedetomidine with remifentanil administration were included. Meta-analyses were performed mostly employing a random effects model. The primary outcome was pain score at rest (visual analogue scale, 0-10) at two postoperative hours. The secondary outcomes included: pain score at rest at 24 postoperative hours; opioid consumption at 2 and 24 postoperative hours; and rates of hypotension, bradycardia, shivering and postoperative nausea and vomiting. Twenty-one randomised trials, including 1309 patients, were identified. Pain scores at rest at two postoperative hours were lower in the dexmedetomidine group, with a mean difference (95%CI) of -0.7 (-1.2 to -0.2), I ²  = 85%, p = 0.004, and a moderate quality of evidence. Secondary pain outcomes were also significantly better in the dexmedetomidine group. Rates of hypotension, shivering and postoperative nausea and vomiting were at least twice as frequent in patients who received remifentanil. Time to analgesia request was longer, and use of postoperative morphine and rescue analgesia were less, with dexmedetomidine, whereas episodes of bradycardia were similar between groups. There is moderate evidence that intra-operative dexmedetomidine during general anaesthesia improves pain outcomes during the first 24 postoperative hours, when compared with remifentanil, with fewer side effects.