The study showed that patients given 30 hours of CBT experienced less pain than those given more than 55 hours of intensive physical therapy, suggesting that the cheaper therapy may also be more effective.
The study included 223 patients who had been referred to a rehabilitation centre for non-specific chronic low back pain. Most of the patients had received treatment, usually physiotherapy, in the past.
The patients were randomised to receive CBT, active physical therapy, a combination of active physical therapy and CBT, or to remain on the waiting list.
All patients receiving therapy attended the clinic three times a week for 10 weeks.
The active physical therapy sessions each lasted one hour 45 minutes and included building aerobic capacity by training for 30 minutes at 65-80 per cent of maximum heart rate.
CBT was administered by a trained therapist. CBT patients also received problem solving training, administered by a clinical psychologist.
At the end of the study, patients in all the treatment groups were found to have reduced disability levels while disability increased in patients still on the waiting list. Patients given CBT had the greatest reduction in disability.
Patients in all treatment groups experienced similar improvements in mood and ability to carry out daily tasks but those who had either CBT or were in the combined treatment group reported less pain than those in the physical therapy group.
Lead researchers Dr Rob Smeets, from the Rehabilitation Centre Blixembosch and the University of Maastricht, said: 'The combination treatment was not better than the single treatment components. We were surprised by that.'
He added that the study supported CBT over physical treatment for patients with chronic back pain: 'It was better in terms of cost and was equally effective.'
Dr Graham Archard, a GP in Bournemouth, Dorset, and member of the Pain Society, said: 'Lower back pain is a hugely important condition. If you can effectively treat it with CBT, this could have huge resource implications.'