We investigate the relationship between area general practitioner (GP) supply and individual body mass index (BMI) in England. Individual level BMI is regressed against area whole time equivalent GPs per 1000 population plus a large number of individual and area level covariates. We use instrumental variables (area house prices and age weighted capitation) to allow for the endogeneity of GP supply. We find that that a 10% increase in GP supply is associated with a mean reduction in BMI of around 1kg/m(2) (around 4% of mean BMI). The results suggest that reduced list sizes per GP can improve the management of obesity.