The anticipation of groundwater quality under the influence of urban and agricultural expansions is an essential issue in environmental problems. Several different chemical and physical parameters affect groundwater quality for drinking purposes. Therefore, the purpose of this study is to comparatively analyze three different prediction approaches to assess groundwater quality for drinking purposes. One of these approaches is multiple linear regression (MLR), while the others are fuzzy inference systems (FIS), including clustering (Model I), and artificial neural network (ANN) model with FIS, including clustering (Model II). In the assessment approaches, clustering analysis is done with the self-organizing map (SOM) methodology, FIS is applied as Mamdani fuzzy system, and ANNs are implemented as feed-forward neural networks. All results of the prediction approaches were compared with the laboratory results. A total of fourteen different chemical and physical parameters were used as inputs for all methods. The results of this study demonstrated that the Model II method developed by the combination of SOM, FIS, and neural networks can be used as an alternative approach for evaluating groundwater quality for drinking purposes as compared with the MLR method, which is a well-known approach.