Tax revenues are the most important sources of public funding. In other words, the role of tax revenues as the most effective tool of fiscal policy, is very important. Therefore, according to the essential role of tax revenue on governments’ economic situation, and since a prerequisite for effective actions in order to implement appropriate fiscal policies, is to know its determinants, this paper tried to investigate the determinants of tax revenues in Iran during 1350-1392, using a Bayesian Econometric Approach and applying Bayesian Model Averaging. The results show that some variables such as literacy rate, GDP growth, population growth, value added of industry sector, and government expenditure have positive effects, but variables such as exchange rate, GINI index, value added of agricultural sector, oil revenue, and value added of oil sector have negative effects on tax revenues. The policy recommendations resulted from this study are: To increase tax revenues through structural changes in economy by increasing the share of industry sector and industrialization of agricultural sector; To improve tax laws and to find executive effective solutions in order to decrease big investors tax evasion; and to try in order to increase efficiency and effectiveness of the education system.