Seventy-one chronic low back pain patients were assigned to one of three behavioural rehabilitation treatments or a waiting-list condition. The first intervention consisted of an operant treatment, aimed at increasing health behaviours and activity levels and at reducing pain and illness behaviours. In the second intervention, a cognitive treatment, aimed at the reinterpretation of catastrophizing pain cognitions and at enhancing self-control, was combined with an operant treatment. The third intervention consisted of the combination of the operant approach and a respondent treatment. During the respondent treatment, patients were taught to decrease muscle tension levels, using the 'applied relaxation' technique supported by EMG-biofeedback and graded exposure to tension-eliciting situations. A repeated measurements design included observer rating of pain behaviours, observer ratings of mood, self-reported depression, residual health behaviours, pain cognitions and experienced pain intensity. Follow-up assessment occurred at six months and one year after termination of treatment. Results suggest that, for the sample as a whole, improvements are found on measures of pain behaviours, health behaviours, pain cognitions and affective distress and that these improvements are maintained at six months and one year follow-up. During the treatment the three treatment groups improved significantly more than the waiting-list control group on most of the measures. Further, the results of this study provide evidence that the operant-cognitive and operant-respondent conditions are more efficacious in decreasing pain behaviours and in increasing health behaviours and efficacy expectations than operant treatment alone. This differential effect among the conditions is maintained at follow-up. Patients who received the OC and OR treatments catastrophize less than OP patients, and OC patients showed better scores on outcome-efficacy than OR patients. In general, the results suggest that behavioural rehabilitation programmes for chronic low back pain are effective and that the effects of an operant treatment are magnified when self-control techniques are added.