In this paper we examine the impact of a tailored health warning on health outcomes. We exploit the design of a household panel survey that provided feedback to participants on their blood-pressure levels as a quasi-experiment. We find that many participants who were told their blood-pressure was high went on to get a formal diagnosis of hypertension from a medical practitioner. The effect of getting a formal hypertension diagnosis was to reduce the incidence of smoking and improve the quality of diets. However, we do not find changes in monthly alcohol spending. The behavioural changes (plus any prescribed medications) were large enough to reduce the incidence of cardiovascular disease.