This study examined the efficacy of Internet-delivered cognitive and behavioural interventions for adults with chronic pain, and explored the role of clinical and study characteristics as moderators of treatment effects. PubMed, Embase, PsycINFO, and CENTRAL and CINAHL were searched to identify randomised controlled trials published up to October 2021. A meta-analysis of 36 studies (5778 participants) was conducted, which found small effect sizes for interference/disability (Hedges' g = 0.28; 95% CI 0.21, 0.35), depression (g = 0.43; 95% CI 0.33, 0.54), anxiety (g = 0.32; 95% CI 0.24, 0.40), pain intensity (g = 0.27; 95% CI 0.21, 0.33), self-efficacy (g = 0.39; 95% CI 0.27, 0.52) and pain catastrophizing (g = 0.31; 95% CI 0.22, 0.39). Moderator analyses found interventions which involved clinician guidance had significantly greater effect sizes for interference/disability (g = 0.38), anxiety (g = 0.39), and pain intensity (g = 0.33) compared to those without (g = 0.16, g = 0.18; g = 0.20, respectively). Studies using an inactive control had greater effects for depression (g = 0.46) compared to active control trials (g = 0.22). No differences were found between treatments based on traditional Cognitive Behaviour Therapy versus Acceptance and Commitment Therapy. Sample size, study year, and overall risk of bias (Cochrane rating) did not consistently moderate treatment effects. Overall, the results support the use of internet-delivered cognitive and behavioural interventions as efficacious and suggest guided interventions are associated with greater clinical gains for several key pain management outcomes.Prospectively registered on OSF Registries (citation: osf.io/cvq3j/).