The management of back pain is controversial. Despite a high incidence (between 15% and 40% in most Western countries) and associated economic burden, considerable uncertainty exists as to the effectiveness and cost-effectiveness of alternative interventions for the condition. OBJECTIVES: To determine whether, from a UK National Health Service (NHS) perspective, surgical stabilisation of the spine is cost-effective when compared to an intensive rehabilitation programme for the treatment of patients with chronic low back pain. STUDY DESIGN: Three hundred forty-nine patients (349) assessed as having chronic low back pain were randomised to surgery (176 patients) or rehabilitation (173 patients) at centres across the UK. Patients were followed-up at 6, 12, and 24 months post randomisation. METHODS: Costs to the NHS of initial treatment (surgery or rehabilitation), medications, and all primary and secondary sector health care contacts were collected for each patient out to 24 months. Patient utility measured using the EuroQol EQ-5D questionnaire was combined with 24 month survival data to calculate quality adjusted life years (QALYs). Results were expressed using an incremental cost per QALY. Statistical techniques were used to examine stochastic uncertainty surrounding cost, QALY, and cost per QALY results. RESULTS: Preliminary analysis shows surgery to be more costly than rehabilitation at 24 months following randomisation. The main cost drivers appear to be the initial surgical procedure and a higher proportion of surgery patients receiving subsequent outpatient and community care. No difference in QALYs was detected between the two modes of treatment and the baseline incremental cost per QALY exceeded £30,000. Examination of uncertainty surrounding key parameters did not alter these results greatly. CONCLUSION: Preliminary results from this trial, one of very few in orthopaedic surgery in the UK, suggest that surgical stabilisation of the spine for patients with chronic low back pain may be more costly than alternative treatments with no clear advantage in successful outcomes.