The present work is focused on ship’s parametric roll. In particular, a series of controlled experiments have been performed in order to obtain repeatable results both for the inception and the roll amplitude in parametric roll condition. The experiments have been performed for an innovative ship design (a tumblehome shape) in a towing tank provided by a flap-type wave maker. The model was tested in 3 DoF conditions, being roll, pitch and heave motions free and surge, sway and yaw restrained. The experimental work, which is the subject of the present paper, is part of cooperative research project between the IIHR (Iowa Institute of Hydraulic Research), the University of Trieste and the Italian ship model basin (INSEAN) (Olivieri et al, 2006a, Olivieri et al, 2006b). All the experiments have been carried out in head waves at the INSEAN basin n. 2. The model motions were measured both by an optical motion tracker (Krypton) and a gyroscopic platform (MOTAN). Tests were carried out for three different metacentric heights in a range of speeds. Roll decay tests as a function of forward speed were performed before every experiment in waves in order to determine the natural roll frequency and the roll decay rate. The head waves experiments showed parametric roll excitation for different metacentric heights, wave steepness and Froude number. The repeatability of the obtained results, allows to use them as a benchmark for both CFD codes and theoretical models (Sadat-Hosseini et al, 2010). In the present paper the focus is on the experimental part of the work including uncertainty analysis based both on repeatability and non-random uncertainty sources. In particular, the uncertainty on the motion (roll) measurements has been determined by using contemporaneously two different devices (Krypton and MOTAN), taking as a measure of the bias the difference between the two outcomes. The statistical part of the uncertainty has been determined by repeating the tests 5 times for a set of given conditions.