Sills are the initial foundation or lower part of the structures where water jets falls from the hydraulic structure or any natural structure. Sills may also be used as part of scheme to maintain a minimum water level in rivers and in estuary. The scour phenomena occur when plunging jets hits the bed sills causing displacement of bed particles. The prediction of scour depth is essential to determine the stability of the structures around it. Scouring effect often ignored by people due to it unseen able characteristic but it can cause massive damage if been overlooked. The study of the scour downstream of sill has been conducted using regression equation but there are weaknesses regarding the best possible way to predict scour depth. Thus in this study, multiple linear regression, adaptive neuro-fuzzy inference system (ANFIS), Feed Forward Back Propagation (FFBP) neural network has all been used to estimate the scour depth. Laboratory data from an ICE journal has been adopted for this project. The validation of the developed network has been done by using observed data which were not involved in training process.