Current guidelines recommend opioid therapy to chronic non-malignant pain (CNP) patients when the benefits for pain and function outweigh risks. This systematic review examined the effects of opioid therapy on sleep - a valued functional outcome- in CNP. Electronic and hand searches of relevant studies up through July 2017 identified 18 eligible studies providing data from 3,746 CNP patients for analysis. Twelve of these studies were randomised controlled trials of up to 12-month in duration. Morphine sulfate, oxycodone and transdermal fentanyl were the most tested therapies (n = 4 each). Only two studies used objective sleep measures in addition to self-report ratings, questionnaires or sleep diaries. Whilst calmer sleep with less body/leg movements and fewer awakenings could be achieved following opioid therapy, these might occur with increased sleep-disordered breathing and a much-shortened rapid eye movement (REM) sleep latency. Both the narrative synthesis and exploratory meta-analysis suggest that opioid therapy in CNP is associated with improved self-reported sleep quality. However, the effect is inconsistent, small (Standardised Mean Difference = 0.36), and may be accompanied by excessive daytime sleepiness. As a Cochrane-recommended assessment revealed "unclear" or "high" overall risk of bias for all studies, future opioid trials of stronger methodology and better reporting are needed to confirm and elucidate the effect.