Studies on the ship behaviour in waves have recently improved so much by the statistical approach, so that we might foresee some quantitative results about the ship behaviour even at the stage of initial design. Sofaras our knowledge is concerned, the data about the destroyer which are available at the initial design stage, are considered very few. In order to examine the motion of destroyer in rough seas and guess previously some motion of them, the sea tests which covering different sea states were carried out using two sister ships. The items measured are pitching and rolling angles, acceleration, heaving motion, hull girder stress, local stress on the bottom panel and slamming pressure on the bottom of forebody ; and all of these recorded data are used for statistical and spectral analyses by applying time series procedure. Theoretical calculations, on the other hand, under the cross flow hyposesis are pursued in order to obtain the response amplitude operators of pitching angle, heaving motion, bending moment and relative motion etc., then the power spectra of them are also calculated by estimating wave spectra of the experimental water areas. Comparing with both calculated and analysed results, we obtain the facts which distributions of variation in pitching angle, heaving motion and hull girder stress etc. follow to Rayleigh's law, allowing from 67% to 90% confidence limit, and slamming pressures to truncated exponential distribution ; and also from the power spectra the pitching angle, heaving motion and acceleration show good coincidences from the practcal view-point, but on the contrary bending moments some discrepancies especially at the point of maximum magnitudes. Number of occurrences of slamming impact and the distribution of slamming pressure are theoretically calculated under several assumptions about the threshold velocity and the bottom shape, and their results are compared with recorded data showing some discrepancies for number of occurrences, but for the distribution of slamming pressure good coincidences.