Planning of hydroelectric systems is a complex and difficult task once it involves non-linear production characteristics and depends on numerous variables. A key variable is the streamflow. Streamflow values covering the entire planning period must be accurately forecasted because they strongly influence energy production. This paper suggests an application of a FIR neural network and a fuzzy clustering-based model to evaluate one-step and multi-step ahead predictions. Results are compared to the ones obtained by a periodic autoregressive model (PAR). It is interesting to apply a recurrent neural network for prediction task due to its ability for temporal processing and efficiency to solve nonlinear problems. The results show a generally better performance of the FIR neural network for the case studied. © 2005 IEEE.4 26312636Maier, H., Dandy, G., Neutal networks for the prediction and forecasting of water resources variables: A review of modelling issues and applications (2000) Environmental Modelling & Software, 15, pp. 101-124Box, G., Jenkins, G., Reinsel, G.C., (1994) Time Series Analysis, Forecasting and Control, 3rd Ed., , Oakland, California, EUA: Holden DayWeigend, A.S., Gershenfeld, N.A., (1992) Time Series Prediction: Forecasting the Future and Understanding the PastSlim, C., Trabelsi, A., Neural network for modeling financial time series: A new approach (2003) ICCSA, (3), pp. 236-245Lapedes, A., Farber, R., Nonlinear signal processing using neural networks: Prediction and system modelling (1987) Tech. Rep. LA-UR-&-2662, , Los Alamos National LaboratoryKim, H.J., Lee, W.D., Yang, H.S., A modified FIR network for time series prediction (2002) Proceedings of the 9th International Conference on Neural Information Processing, 5, pp. 2597-2600Ku, C., Lee, K., Diagonal recurrent neural networks for dynamic systems control (1995) IEEE Transactions on Neural Networks, 6 (1), pp. 144-156Oh, K., Han, I., An intelligent clustering forecasting system based on change-point detection and artificial neural networks: Application to financial economics (2001) Proceedings of the 34th Hawaii International Conference on System ScienceGeva, A., Non-stationary time series prediction using fuzzy clustering (1999) Proceedings of the International Conference of the North American Fuzzy Information Processing Society, pp. 413-417Wan, E., Temporal backpropagation for FIR neural networks (1990) International Joint Conference on Neural Networks, 1, pp. 575-580. , JuneMagalhães, M., Ballini, R., Gonçalves, R., Gomide, F., Predictive fuzzy clustering model for natural streamflow forecasting (2004) Proceedings of the IEEE International Conference on Fuzzy Systems, pp. 390-394. , Budapest, Hungary, JulyHaykin, S., (2001) Neural Networks: A Comprehensive Foundation, 2nd Ed., , Prentice Hall, IncWang, L.-F., Li, X.-X., Model identification of time delay nonlinear system with FIR neural network (2003) Proceedings of the Second International Conference on Machine Learning and Cybernetics, 2, pp. 872-875. , NovemberBezdek, J., (1981) Pattern Recognition with Fuzzy Objective Function Algorithms, , New York, EUA: Plenum PressSchwarz, G., Estimating the dimension of a model (1978) The Annual of Statistics, 6 (2), pp. 461-464