The present study investigated the ability of sociodemographic variables to predict the short-term effect of multidisciplinary pain treatment in 286 chronic non-malignant pain patients consecutively referred to a Danish multidisciplinary pain centre. At inclusion and 3 and 6 months later the patients' pain and health related quality of life (HRQL) was assessed using self-administered questionnaires. Sociodemographic variables measured were: age, gender, educational level, civil status, employment status and disability pension (DP) status. Intensity of pain was measured using a VAS scale. HRQL was evaluated using the Medical Outcome Study-Short Form (SF-36) and the Psychological General Well-Being Scale (PGWB). Drop out was 30% after 6 months, leaving 170 patients for the study of treatment effects. Among the sociodemographic variables evaluated only retirement pension (RP) and DP-status seemed to be significant outcome predictors: six months after inclusion, patients on RP (patients >68 years, n = 40) reported larger pain reduction (30 mm on the VAS scale) than patients < or =68 years (8 mm on the VAS scale)(P = 0.001). Patients applying for DP-pension (n = 26) did not improve. Patients on DP (n = 54) and patients not receiving DP (n = 50) obtained moderate improvements, but the improvements were significantly larger than the improvements observed in the ApplyDP-group (P = 0.05). The same pattern was seen for changes in psychological well-being (PGWB) and social functioning (SF36-SFA). No improvements were obtained in SF36-Physical functioning scores. DP-status predicted improvement in pain and social functioning even when controlling for the effect of other demographic variables, pain intensity and HRQL at referral. The present study indicates that the multidimensional problems experienced by patients applying for DP are dominated by socioeconomic factors. Focus on solving these socioeconomic problems is important if chronic pain patients are to benefit from multidisciplinary pain treatment.