This systematic review and meta-analysis critically examined the evidence for peer support interventions to reduce pain and improve health outcomes in community-dwelling adults with chronic musculoskeletal pain (PROSPERO CRD42022356850). A systematic search (inception-January 2023) of electronic databases and grey literature was undertaken to identify relevant randomised controlled trials, with risk of bias and GRADE assessments performed on included studies. Meta-analyses used a generic, inverse-variance, random-effects model, calculating mean difference (MD) or standardised mean difference (SMD). Of 16,445 records identified, 29 records reporting on 24 studies (n = 6202 participants) were included. All evidence had unclear/high risk of bias and low-very low certainty. Peer support interventions resulted in small improvements in pain (medium-term: MD -3.48, 95% CI -6.61, -0.35; long-term: MD -1.97, 95% CI -3.53, -0.42), self-efficacy (medium-term: SMD 0.26, 95% CI 0.16, 0.36; long-term: SMD 0.21, 95% CI 0.07, 0.36), and function (long-term: SMD -0.10, 95% CI -0.19, -0.00) relative to usual care and greater self-efficacy (medium-term: SMD 0.36, 95% CI 0.20, 0.51) relative to waitlist control. Peer support interventions resulted in similar improvement as active (health professional led) interventions bar long-term self-efficacy (MD -0.41, 95% CI -0.77, -0.05), which favoured active interventions. No point estimates reached minimal clinically important difference thresholds. Pooled health service utilisation outcomes showed unclear estimates. Self-management, quality of life, and social support outcomes had mixed evidence. Despite low-very low evidence certainty, peer support interventions demonstrated small improvements over usual care and waitlist controls for some clinical outcomes, suggesting that peer support may be useful as an adjunct to other treatments for musculoskeletal pain.