OBJECTIVE:To explore whether psychosocial or demographic factors are associated with early dropout from pain self-management in a rural, low-socioeconomic status population. DESIGN:Secondary analysis of retrospective data. SETTING:Multidisciplinary pain clinic located in an outer regional area of Australia. SUBJECTS:One hundred eighty-six people attending a public community health center with chronic noncancer pain (mean age 54.9 years; 58.1% women; 81.7% in receipt of government benefit as their primary source of income). METHODS:Bivariate analysis and logistic regression, with early dropout as the dependent variable and a range of demographic and psychological independent variables. RESULTS:Following bivariate analysis, early dropout was significantly associated (P < 0.05) with male gender, younger age, history of substance use, being a past victim of assault/abuse, receiving unemployment or disability benefit, having literacy difficulties, higher pain catastrophizing score, higher daily opioid dose, and not holding a multifactorial belief about the cause of pain. Logistic regression analysis resulted in three significant predictors of dropout: substance use history (P = 0.002), past victim of assault or abuse (P = 0.029), high pain catastrophising score (P = 0.048); and one of engagement: holding a multifactorial belief about pain cause (P = 0.005). CONCLUSIONS:In a rural, low-socioeconomic status population, addressing social stressors related to lifetime adversity may be important to increasing engagement in pain self-management. Lack of attention to these factors may increase health inequity among those most disabled by chronic pain. Further research into dropout and engagement, especially among disadvantaged populations, is recommended.