Monitoring procedures of groundwater quality parameters around mines and landfills are facing two major problems: First, they are labor intensive and expensive; and second, the sampling data are sparse and include too many parameters. Therefore, the site characterization schemes should be optimized. This research is mainly concerned with groundwater resources characterization. The data involved are either continuously monitored multi-parametric water quality measurements from piezometric wells or come from direct push penetrometric sampling. The data collected are spatially referenced since the geographical XY co-ordinates of the sampling point are known. Therefore, for each variable measured in situ, the data have a 3D reference. A methodology has been developed for 3D characterization of groundwater resources using multivariate statistics methods and geostatistics techniques. The developed system enhances the quality and spatial distribution of the measured groundwater data and enables the prediction of quality parameters. The methodology could be a useful tool for assisting a water quality-monitoring programme by providing a quick and reasonably accurate means of data estimatio