Fifty-eight outpatients with chronic low back pain were randomly allocated to one of six experimental conditions. Four conditions were designated as treatment conditions and two as control conditions. The four treatment groups consisted of: cognitive treatment (either with or without relaxation training) and behavioural treatment (either with or without relaxation training). The cognitive and behavioural groups also received physiotherapy. The two control conditions consisted of: attention (physiotherapy plus discussion sessions) and no-attention (physiotherapy-only) conditions. All conditions, including the two controls, received the same physiotherapy back-education and exercise program. For the sample as a whole, improvements were obtained on measures of affective distress, functional impairment, medication use, pain-related dysfunctional cognitions and use of active coping strategies. These improvements were generally maintained at 6- and 12-month follow-ups. The combined psychological/physiotherapy treatment conditions improved significantly more than the physiotherapy-only conditions from pre to posttreatment on measures of pain intensity, self-rated functional impairment and pain-related dysfunctional cognitions. However, these differences were only weakly maintained at 6- and 12-month follow-ups. The behavioural conditions improved significantly more than the cognitive conditions from pre to posttreatment on the self-rated measure of functional impairment, but this difference was not maintained at 6- and 12-month follow-ups. Progressive relaxation training was found to make little contribution to either cognitive or behavioural treatments.