This thesis is concerned with the usage of models for prediction of groundwater flow and quality in aquifers. A critical review of current practical applications of groundwater models and modelling trends has been undertaken. On the basis of this study and taking into account the recent increase in the availability of relatively cheap microcomputers, several potential lines of alternative modelling techniques have been examined. The uncertainty inherent in the collection of any groundwater data from the field and the identification of boundary conditions suggest that the commonly employed large complex distributed-parameter models are no more valuable as predictive tools than less complex alternatives, which have been largely neglected. The recent rapid developments in microcomputer technology and availability provide an opportunity for much wider use of groundwater models in planning and management as well as the more traditional scientific and engineering applications. Three prototype models, utilising new or neglected techniques and taking advantage of the wide availability of microcomputers, are presented. These models seem to offer new possibilities in modelling to a broad range of those involved in practical groundwater management and resource planning.