Conversion of rainfall to runoff according to the laws of gravity vivifies earth, replenishes groundwater, keeps rivers and lakes full of water, and varies the landscape by the action of erosion. Large uncertainties and high non-linearity of the Rainfall-Runoff (R-R) process make it complex task to have the process-based modeling, so it is preferred to create a black box relationship between driving and resultant variables. Therefore, several black box approaches including Artificial Intelligence (AI) models have been developed and used to simulate R-R process. In this paper, WANN and WSVM models are employed for Multi-Station (MS) modeling of R-R. However in any data driven modeling, some of the inputs may have no significant relationship with the output variables. Therefore, determination of dominant input variables, is one of the indispensable challenges in the model development procedure. For this purpose, to extract main features and inputs of the WANN, WSVM methods, two kinds of data pre-processing methods of Self-Organizing Map (SOM) based clustering and Mutual Information (MI) concepts are employed in this study. Therefore, spatio-temporal investigation, identification and using all sub-basins records as a cascade-based MS analysis can improve prediction of runoff in watersheds. For this purpose, two scenarios with distinct views were used for MS modeling of R-R to identify the suitable strategy for future hydro-environmental researches. The results indicated that the proposed AI-models coupled with the SOM and MI tools improved the performance of MS runoff prediction compared to the Markovian-based models up to 23%. Nevertheless, benefit of the seasonality of the process along with reduction of dimension of the inputs could help the AI-models to consume pure information of the recorded data.