Background: The underlying processes of change that contribute to the effectiveness of multidisciplinary pain treatment require clarification. Previous research has found support for pain acceptance as a process variable in acceptance-based treatment. Preliminary findings indicate that pain acceptance may also be a process variable in traditional cognitive behavioural therapy (CBT). The aim of this study was to investigate the role of pain acceptance as a process variable in CBT relative to two empirically supported process variables, namely catastrophizing and pain intensity. Methods: Patients with chronic pain (n = 186) attended a 3-week, multidisciplinary pain programme, which was CBT based. Patients completed a measure of pain intensity; the Chronic Pain Acceptance Questionnaire; the catastrophizing subscale of the Pain Response Self-Statements Scale; the Roland Morris Disability Questionnaire; the Depression Anxiety and Stress Scale; and two measures of physical functioning at pretreatment, post-treatment and 3-month follow-up. Results: Both acceptance and catastrophizing showed statistically significant and clinically relevant changes from pre- to post-treatment. Changes in both acceptance and catastrophizing showed a significant correlation with changes in almost all of the outcome variables. Regression analyses demonstrated that change in acceptance was a significant predictor of changes in depression, disability, timed walk and sit-to-stand performance, after controlling for changes in catastrophizing and pain intensity. Conclusions: Although not specifically targeted in CBT treatment, acceptance of pain was an important process variable that contributed to CBT treatment outcomes after controlling for changes in pain intensity and catastrophizing. Implications for future research and clinical practice are discussed.