The main target of this research is providing a decision making model to identify the target groups for energy subsidy reform using artificial neural network approach. Family expenditures on electricity, gas, telephone and cell phone are used as the inputs to the artificial neural network model and the Probit model developed as a benchmark. The sample being studied in this research are divided into two groups: eligible and non-eligible to receive subsidies. The artificial neural network used in the research is the multilayer perceptron which has used the Levenberg-Marquardt method to train the data.Acooring to the results, there exists a statistically meaningful relation between the families’ expenditure on the chosen services and their belonging to different income levels as eligible or non-eligible groups for receiving subsidies. Moreover, families with higher expenditures on selected services have higher incomes.The outcome of this study reveals that the two models used for classifying the target groups have similar results in classifying the families into eligible and non-eligible groups for receiving subsidies.