Chronic low back pain is a common, disabling and costly health problem. The treatment of chronic low back is difficult and is often ineffective. For treatment to be effective the cause of the pain has to be established but unfortunately in 80% to 95% of the patients the cause cannot be determined despite the existence of modern imaging techniques. A pathoanatomical diagnosis which fits into a classical disease model where successful treatment can be carried out, can only be made in 5% to 7% of the patients. The back pain in the rest of the patients where no pathoanatomical diagnosis can be made is often labelled, unscientifically, as chronic low back pain. Despite the existence of sophisticated imaging techniques and a plethora of diagnostic test the source of pain in patients with nonspecific back pain cannot be established. There exist no causal relationship between imaging findings of degenerated disc, lumbar facet arthritis, spondylosis, spondylolysis and spondylolisthesis, to the pain in these patients. Surgical treatment of non-specific back pain where no pathoanatomical diagnosis has been established is bound to fail. Therefore the outcome of spinal fusion in these patients can be no better than nonsurgical treatment. Spinal fusion is a major surgery which can be associated with significant morbidity and occasionally with mortality. Yet there is rapid rise in the rates of spinal fusion. There is a growing tension between ethics and conflicts of interest for surgeons. The spine, unfortunately, has been labelled as a profit centre and there are allegations of conflicts of interest in the relationship of doctors with the multi-billion dollar spinal devices industry. The devices industry has a significant influence on not only research publications in peer review journals but also on decisions made by doctors which can have a detrimental effect on the welfare of the patient.