To determine the amount of dependable flow, a hydrological approach is needed where changes in rainfall become runoff. This diversification is a very complex hydrological phenomenon. Where this is a nonlinear process, with time changing and distributed separately. To approach this phenomenon, an analysis of the hydrological system has been developed using a model which is a simplification of the actual natural variables. The model is formed by a set of mathematical equations that reflect the behavior of parameters in hydrology. Modeling in this case uses artificial neural networks, multilayer perceptron combined with the backpropagation method is used to study the rainfall-runoff relationship and verify the model statistically based on the mean square error (MSE), Nash-Sutcliffe Efficiency (NSE) and correlation coefficient value (R2). Of the three models formed, model 3 provides optimum results with correlation levels using NSE per month as follows, in Cikapundung Sub-Basin NSE = 0,990703, R2 = 0,995008, and MSE = 0,00014443, while in Citarik Sub-Basin NSE = 0.9500, R2 = 0.97592, and MSE = 0.0010804 . From these results it can be seen that ANN has a fairly good ability to replicate random discharge fluctuations in the form of artificial models that have almost the same fluctuations and can also be applied in rainfall runoff modelization even though the results of the test results are not very accurate because there are still irregularities