Many researches were accomplished to predict the maximum scour depth around piers located in a river main channel. Based on the results of previous studies many equations were proposed for this purpose, however, each equation incorporates only limited affecting facorts on the phenomena. In this study valuable data gathered during the last few decades have been used to examine the application of Artificial Neural Networks for predicting the maximum local scour depth around piers. The data were randomly divided into two parts, the first part was used to train the ANN and the second to validate its outputs. This research indicated that reasonable concordance was obtained between observed local scour depths and calculated values based on ANN. ANN also, yields more satisfactory results as compared with the multi linear regression model. Sensitivity analysis was also done to determine the most affecting factors on the process. It was indicated that the geometric standard deviation of the bed material affects the output the most. Other factors such as flow velocity, median grain size, pier width, and flow depth affect the phenomenon as well.
)