The paper presents the application of a backanalysis methodology to the inverse analysis of preloading soft clays. The methodology is based on an appropriate selection of the objective function and it uses a novel adaptive genetic algorithm approach to solve the resulting optimization problem. The methodology is applied first to a synthetic example followed by an application to an intensely instrumented preload test. The selected observations are the evolution along time of displacements and pore water pressures. The key parameters to be identified are those with major influence on the results: compressibility and permeability of the soil. It is shown that the confidence in parameter estimation varies significantly depending on the time at which the observations are considered and the type of them (displacements vs. pore water pressure). The possibility of extrapolating early results to later stages of the preloading is also explored. The usefulness of the method in engineering problems is demonstrated by the application to a preloading performed to improve the soft ground of the Water Treatment Plant located in the Llobregat delta near Barcelona.