In this paper, three econometric models (linear regression model, Reciprocal model & Double-log or log-linear regression model) comprised one outcome variable (infant mortality rate) and three explanatory variables (per capita gross domestic product, per capita health expenditure and mean age at marriage of female) have been considered over the period 1980 to 2007 for Bangladesh. At all stages of fitting model, it has been observed that at least one of the independent variables had significant effect on infant mortality rate. After removing multicollinearity and autocorrelation, two independent variables (per capita health expenditure and mean age at marriage of female) had significant effect on infant mortality rate in case of reciprocal model but per capita health expenditure had substantial effect on infant mortality rate in three econometric models. Results indicate for the policy makers at the national level for raising per capita health expenditure to decrease infant mortality rate.