Modern statistical inference makes use of two major paradigms: the frequentist and the Bayesian ones. In both cases the goal is to estimate parameters from observations, by solving an inverse problem. But while in a frequentist approach the unknown parameters are usually considered as fixed values, not allowing to account for any a priori information, in the Bayesian methods they are treated as random variables to which a probability or a probability density function (if continuous) is associated. Therefore the Bayesian approach can be efficiently used to solve a large number of problems. Prof. Fernando Sansò was one of the first to grasp the potentiality of this method for the geodetic disciplines. He and many people from his group have been working on this subject since a long time. In the following we will present few examples of geodetic applications which Fernando Sansò worked on.