The efficacy of music for the treatment of pain has not been established.To evaluate the effect of music on acute, chronic or cancer pain intensity, pain relief, and analgesic requirements.We searched The Cochrane Library, MEDLINE, EMBASE, PsycINFO, LILACS and the references in retrieved manuscripts. There was no language restriction.We included randomized controlled trials that evaluated the effect of music on any type of pain in children or adults. We excluded trials that reported results of concurrent non-pharmacological therapies.Data was extracted by two independent review authors. We calculated the mean difference in pain intensity levels, percentage of patients with at least 50% pain relief, and opioid requirements. We converted opioid consumption to morphine equivalents. To explore heterogeneity, studies that evaluated adults, children, acute, chronic, malignant, labor, procedural, or experimental pain were evaluated separately, as well as those studies in which patients chose the type of music.Fifty-one studies involving 1867 subjects exposed to music and 1796 controls met inclusion criteria. In the 31 studies evaluating mean pain intensity there was a considerable variation in the effect of music, indicating statistical heterogeneity ( I(2) = 85.3%). After grouping the studies according to the pain model, this heterogeneity remained, with the exception of the studies that evaluated acute postoperative pain. In this last group, patients exposed to music had pain intensity that was 0.5 units lower on a zero to ten scale than unexposed subjects (95% CI: -0.9 to -0.2). Studies that permitted patients to select the music did not reveal a benefit from music; the decline in pain intensity was 0.2 units, 95% CI (-0.7 to 0.2). Four studies reported the proportion of subjects with at least 50% pain relief; subjects exposed to music had a 70% higher likelihood of having pain relief than unexposed subjects (95% CI: 1.21 to 2.37). NNT = 5 (95% CI: 4 to 13). Three studies evaluated opioid requirements two hours after surgery: subjects exposed to music required 1.0 mg (18.4%) less morphine (95% CI: -2.0 to -0.2) than unexposed subjects. Five studies assessed requirements 24 hours after surgery: the music group required 5.7 mg (15.4%) less morphine than the unexposed group (95% CI: -8.8 to -2.6). Five studies evaluated requirements during painful procedures: the difference in requirements showed a trend towards favoring the music group (-0.7 mg, 95% CI: -1.8 to 0.4).Listening to music reduces pain intensity levels and opioid requirements, but the magnitude of these benefits is small and, therefore, its clinical importance unclear.