This study aimed to investigate the relationships among Gross Domestic Product (GDP) per capita, Government Health Expenditure (GHE) per capita and Infant Mortality Rate (IMR) in China, and explore the new model to reduce the IMR. A periodic bivariate Pearson’s analysis was performed on the data of GDP per capita, GHE per capita and IMR in China from 1952 to 2014. The turning point for the effect of GDP per capita on IMR was sought. Then, the average development levels of these three indicators were compared. Finally, the decline characteristics of IMR were analysed. Results showed that, there was no relationship between GDP per capita and IMR from 1952 to 1970 (r=-0.393, p=0.096). GDP per capita and IMR were negatively correlated from 1970 to 2014 (r=-0.827, p=0.000). The stage from 1970 to 1975 was the initial turning point for the effect of GDP per capita on IMR (r=-0.980, p=0.001). From 1980 to 1995, the growth rate of GHE per capita (12.80%) was slower than that of GDP (17.26%), and the average decline rate of IMR was 2.82%. From 1995 to 2014, the average growth rate of GHE per capita (18.25%) was faster than GDP per capita (12.42%), and the average decline rate of IMR was significantly accelerated (7.15%). The negative correlation of GDP per capita and IMR does not always exist. The effect of GHE per capita on IMR is more obvious than that of GDP per capita. It is suggested to build a “surpass-growth and support” mode to reduce the IMR.