Most high-income older Australians became eligible for a pharmaceutical concession through a 1999 policy change. I exploit this natural experiment to estimate their price elasticity of demand for pharmaceuticals. The preferred model is a nonlinear Instrumental Variable (IV) regression, estimated on nationally representative repeated cross-sectional survey data using the Generalized Method of Moments (GMM). No significant evidence is found for endogenous take-up, and so cross-sectional estimates are also considered. Taking all of the results into account, the ‘headline’ estimate is −0.1, implying that quantity demanded is not highly responsive to price.