Music is increasingly used as an adjuvant for chronic pain management as it is not invasive, inexpensive, and patients usually report positive experiences with it. However, little is known about its clinical efficacy in chronic pain patients. In this systematic review and meta-analysis, we investigated randomized controlled trials (RCTs) of adult patients that reported any type of music intervention for chronic pain, chosen by the researcher or patient, lasting for any duration. Searches were performed using PsycINFO, Scopus and PubMed for RTCs published until the end of May 2016. The primary outcome was reduction in self-reported pain using a standardized pain measurement instrument reported post-intervention. The secondary outcomes were: quality of life measures, depression and anxiety measures, among others. The study was pre-registered with PROSPERO (CRD42016039837) and the meta-analysis was done using RevMan. We identified 768 titles and abstracts, and we included 14 RTCs that fulfilled our criteria. The sample size of the studies varied between 25 and 200 participants. We found that music reduced chronic pain, and depression, with higher effect size on pain and depression. We also found music had a higher effect when the participant chose the music in contrast with researcher-chosen music. The sample size of RCTs was small and sometimes with different outcome measures. Our analysis suggests that music may be beneficial as an adjuvant for chronic pain patients, as it reduces self-reported pain and its common co-morbidities. Importantly, the analgesic effect of music appears higher with self-chosen over researcher-chosen music.