The objective of this study is to analyze impact of various types of government expenditures on poverty reduction in Iran. Estimation is based on time series data 1981-2012, using a system of equations including variables affecting poverty. The system equations include poverty, labor productivity, employment, wage, access to facilities (Electricity, pipe water, and telephone), educational attainment, and life expectancy that according to simultaneity and diagonality of variance –covariance matrix of residual tests has been estimated with “three stage least square” method. Different types of public spending may have differential effects on poverty reduction. Therefore, government expenditures are disaggregated into different types: social affairs (education, health, social security, housing, rural community development and urban community development) and economic affairs (including roads, telecommunications, water, roads, and electricity). Marginal effects of public spending on poverty are calculated. The results reveal that public investments on health, education, social security, and economic infrastructure (including roads and telecommunications), still have positive and significant marginal impacts on poverty reduction and totally spending on social sectors is more likely to benefit the poor than are other types of expenditure.