Objective: To determine the usage of surgery in the management of obesity. Design: Analysis of routine statistics. Data sources: Hospital Episode Statistics extract held by Public Health Observatories. Geographical areas: Two Government Office Regions in England with a population of 7.6 million. Time period: Data analysed in the study covered the 6 y 1996/1997\2001/2002. Results: The availability and uptake of surgery for obesity in the two regions was hugely variable. The overall utilisation of the procedures examined (OPCS operation codes: G30.1, G30.2, G32.10, G61.00) was 5.6 per million per annum, but ranged by Primary Care Trust from zero (the case in eight of the 50 PCTs in the two regions) to 24.0 operations per million per year. The rates of access to this surgery differed over six-fold between the two regions (1.4 operations per year per million population to 7.9 operations per year per million population). The PCTs with the highest rates of surgery, were those closest to the large providers of service. Conclusions: Access to this intervention is highly variable. Primary Care Trusts in England and service providers need to ensure that there is appropriate access to this effective procedure in carefully selected cases. The surgical expertise required for these operations could be concentrated in fewer centres.