Scour is one of the very important problems in the hydraulic engineering, because excessive scour can make the hydraulic structure unstable or decrease its service life. Since scour process is very complicated and many parameters affect it, application of theoretical method seems very difficult so most of the researches in this field are experimental and each researcher has tried to present an equation and estimate scour-hole dimensions. The history of studies in this field shows that most of researchers have addressed estimating maximum depth of scour hole. Whereas in order to control corrosion, all dimensions of scour hole seem important. On the other side in the field of scour-hold depth equations are mostly presented for a small part of data and they do not presented acceptable results based on obtained data from other researches. So in the current research it is tried to present equations and models for estimating all dimensions of the scour hole and in a broader range of data. Data-driven models like neural networks and ANFIS models which are able to predict the behavior of complicated nonlinear functions, nowadays have shown their ability in most fields of science. In this research after analyzing scour process under wall jet effect and investigating effective parameters on it using data collection of different sources, at first it is tried to achieve suitable dimensionless parameters for each dimension of scour hole. Then using the regression method some equations are presented in an available data range. Then adaptive neuro-fuzzy inference system (ANFIS) is developed and the results are compared with the equations of other researchers, the obtained equations of this study and neural network model results. In the end by analyzing the sensitivity, different parameter effects on the hole dimension are investigated. The results of this study show that effective dimensionless parameters on different dimensions of scour hole are different to some extent and also the comparison of models results with the results of obtained equations of statistical methods denotes the superiority of developed models of the current study. Key words: Scour estimation, Submerjed jet, ANFIS, Neural network.