Relationships between life expectancy at birth, public and private health expenditures are analysed with econometric panel time series methods for 34 OECD countries between the years 1970 and 2012. The countries are grouped in three clusters depending on size of public health expenditure as a share of GDP . This gives us the possibility to test whether private and public health expenditures have different impacts on life expectancy at different levels of public expenditure as a share of GDP . Panel unit root tests show that all series in clusters are not difference stationary. Co-integration is found for countries with high public expenditure as a share of GDP . The results, augmented with panel VAR models and impulse response analysis, stress the importance of the positive relationship between public health expenditures and life expectancy. Additional results based on private health expenditure as a share of GDP show also the importance of private expenditures for life expectancy, although these are, at the same time, also driven up by public health expenditures.