Recently, the authors have proposed a new methodology for the vertical position of the center of gravity (KG) estimation of semisubmersibles at its production location. The procedure, called the Zero GM method, is based on the identification of a characteristic behavior of the floater around an induced neutral equilibrium condition. The proposed methodology has been already numerically explored in terms of stability and practical feasibility based on real data of an offshore production semisubmersible. As the procedure implies in temporarily taking the vessel to neutral–unstable upright equilibria (both considered unsafe conditions), experimental tests appear as a physical mean of demonstrating the procedure's practical application and safety. The objective of this paper is to present the results of a model test campaign where the Zero GM method has been experimentally investigated on an offshore semisubmersible. The model and its ballast weight distribution were especially designed to simulate the path of the KG in the process of finding the neutral equilibrium, both in calm-water and in waves. The effect of mooring lines was also assessed. The experimental results showed that around the neutral equilibrium a jump in the loll angle and a peak in the roll oscillation period appear, so that KG can be accurately estimated, even in the presence of (mild) waves and mooring. Capsize or risk of capsize has not been observed during the tests, even when relatively large unstable conditions were tested in waves.