This Ph.D. project aims at development of an improved methodology for soil heterogeneity characterization for ‘next generation’ sampling/monitoring and spatial modeling practices a.o. allowing more realistic pesticide variability in environmental contaminant assessment studies. Such studies typically take place in the laboratory. The key question therefore is: Are current sampling techniques able to counteract the inherent soil heterogeneity met with in the field? Analysis of traditional soil sampling approaches from a Theory of Sampling perspective, the answer is a resounding negative. This contribution summarises the extensive sampling aspects involved in the overall project context, also involving chemometric data analysis with a special twist.