The development and implementation of accurate hydrological forecasting techniques is important for effective and sustainable water resources management, particularly in areas of severe drought. In this study, two artificial intelligence techniques, artificial neural networks (ANN) and wavelet-neural networks (WA-ANN) were developed and compared with the aim of assessing their potential for use in predicting stream flow. The models were developed using data from two rivers in the Mediterranean, a region that is facing severe water shortages and which urgently requires more effective and sustainable water resources management. The two indices for comparison were the coefficient of determination and the root mean squared error (RMSE). The analysis of the developed models showed that both the ANN and WA-ANN methods are suitable potential techniques for hydrological forecasting in this region and for similar watersheds, producing accurate results. In the relative comparison it was shown that the WA-ANN model outperformed the ANN model in both indices for both the rivers. These results indicate that the application of wavelets in conjunction with artificial intelligence methods in hydrological forecasting is a promising new area of research.