This network meta-analysis aimed to estimate the comparative effectiveness of non-pharmacological interventions on sleep in individuals with chronic musculoskeletal pain. Seven databases were systematically searched up to February 2023. A random-effects network meta-analysis in a frequentist framework was performed to synthesize continuous data as standardized mean differences (SMD) along with a 95% confidence interval (95% CI). A total of 15,641 records were identified, and 107 randomized controlled trials involving 8,121 participants were included. Of 14 identified interventions, eight were significantly more effective than passive control in improving sleep quality at immediate post-intervention (SMDs = 0.67–0.74), with cognitive behavioral therapy (CBT) being the most effective treatment (SMD = 0.74, 95% CI: 0.45–1.03). Only CBT demonstrated sustained effects at short-term (SMD = 1.56; 95% CI: 0.62–2.49) and mid-term (SMD = 1.23; 95% CI: 0.44–2.03) follow-ups. Furthermore, CBT significantly improved subjective (SMD = 0.64; 95% CI: 0.25–1.03) and objective (SMD = 0.30; 95% CI: 0.01–0.59) sleep efficiency compared with passive control at immediate post-intervention. Our findings support CBT as the first-line treatment for improving sleep in individuals with chronic musculoskeletal pain, given its superior effectiveness across multiple sleep outcomes and its sustainable effects until mid-term follow-up. However, the certainty of evidence for these interventions in improving sleep quality was very low to low.