A gamma-ray transmission technique is present to measure the void fraction and identify the flow regime of a two-phase flow using two detectors which were optimized in terms of detector orientation. Using Monte-Carlo simulation, experimental results were utilized for training an artificial neural network. Radial Basis Function was used to classify flow regimes (annular, stratified and bubbly) and predict the value of void fraction. All of the training and testing data sets were determined correctly and the mean relative error percentage of predicted void fraction was less than 1.5%. Although the method was applied to a certain pipe size in a static flow configuration, it provides a framework for application to other configurations. Copyright (C) 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.