The aim of the article is to quantify the role of time-to-death in shaping health care costs. By combining a detailed break-down of population-wide data on publicly financed health care expenditure with the population and mortality data we are able to estimate a flexible model of live-cycle healthcare expenditure. We show that introducing time to death into the model enables the health care costs model to fit data perfectly. Our results confirm that the cumulative costs of health care over life-span depend on the age of death to a limited extend. As a result, the effect of rising life expectancy has a minor impact on HCE. However, the transitional effect of the changing age structure of the population should affect the aggregate health care expenditure.