Forecasting streamflow have a significant economic impact, since the forecasting can help in agricultural water management and in providing protection from water shortages and possible flood damage. One of the main research topics related to streamflows is estimating the future flows in a stream. In this study two different forecasting models which are adaptive neuro fuzzy inference system ANFIS and layered recurrent artificial neural networks LRNN were applied to Altinsu Curuh river in eastern Black Sea region at Turkey by using the monthly flow values of the river with different effective hydrological factors from the river basin. A comprehensive comparison between ANFIS model and layered recurrent neural networks was achieved using five important evalution parameters. The performances of both models were found to be comparable. However ANFIS models yielded the best result in estimating and forecasting the river.
