The aim of this project is to validate experimentally the effect of the variability of soil properties on the behavior of typical geotechnical structures. The research consists of two mayor tasks. The first one is related with the study of different alternatives to generate random fields, its advantages and its limitations as well as its applicability to various problems in geotechnical engineering (e.g., shallow foundations, pile foundations, slopes, embankments, etc.). The second activity is the experimental work required to validate the analytical model. It includes the construction and test of artificially fabricated specimens in the centrifuge.