Soil organic carbon (SOC) content plays a key role in soil biological, chemical and physical behavior and knowledge about its state and distribution is essential for the effective and sustainable use of soil. Laboratory measurements of SOC are costly and time consuming and have not the possibility to extend the results to similar areas. Recently, the use of remote sensing data for evaluation of SOC as a simple, rapid, inexpensive and even accurate have attracted the attention of researchers. The aim of this study was to evaluate the efficiency and accuracy of ETM+ satellite images for estimating SOC using artificial neural network and regression models. In this way, the digital number values in different bands of ETM+ satellite images have been used. Both regression and neural networks were used to develop the models between organic carbon measured in the laboratory and satellite data. Finally the accuracy of these models was evaluated with R2 and RMSE indices. The statistical analysis shows that these models was fitted using digital number values in the visible and mid- infrared bands of satellite images have the highest coefficients of determination. Moreover, the neural network models have the higher accuracy and .less error than regression models.