Stream flow prediction is required to provide the information of various problems related to the design and effective operation of river balancing system. The evaluation of natural and technical science over the past centuries has been closely related to experimental studies and modeling of natural resources. Methods to continuously predict water levels at a site along a river are generally its model based. Hydrologist has relied on individual techniques such as determinates, stochastic, conceptual or black box type to model the complex, uncertain rainfall and consecutive water levels. These techniques provide reasonable accuracy in modeling and prediction of stream flow. River Wainganga has been subjected to water level rise during 2004-2005 and, consequently, the low-laying areas along the bank are in undated, giving problems to local inhabitants, irrigation activities and people properly. Another river Bagh has been connecting to the said river the problem of flood arises more. Therefore predicting water levels has started to attract the attention of the researchers. How this local problem get solved or minimized? An attempt has been made to use the conventional method such as Autoregressive model, more deterministic approach through multi-Linear Regression model and Artificial Neural Network which is capable of identifying complex non-linear relationship between input and output data without attempting to reach understanding into the nature of the process. The performances of these approaches were compared and the best possible result amongst them is the key point of this study.