We calculate that the oceanic torque vectors (x, y, z components) acting on the solid Earth using the Parallel Ocean Climate Model (POCM) output data at 3-day intervals for the period 1988–1995. This study examined three distinct types of torque, with different physical mechanisms: pressure, gravitational, and frictional. The pressure torque is further divided into two parts: that due to Earth's ellipticity and that due to the ocean bottom topography. According to resultant time series the ellipticity pressure torque causes the largest effect for the x and y components. The gravitational torque and the ellipticity pressure torque are shown numerically to be exactly proportional by a factor of −0.49. The topographic pressure torque is somewhat smaller in overall amplitude than the ellipticity-induced torque (the sum of the ellipticity pressure torque and gravitational torque). The ocean bottom frictional torque is found to be negligible in all three components. To validate the torque calculations, we compare the calculated oceanic torques with effective torque derived from the ocean angular momentum (OAM) variation, both from POCM. It is shown that the effective torque associated with the mass term of the OAM agrees remarkably well with the ellipticity-induced torque as required by theory. For the motion term the effective torque can be explained mostly by the topographic pressure torque in the intraseasonal periods, but the seasonal and longer-period components show significant differences between the two. The difference should theoretically be attributed to the exclusion of the wind frictional torque on the ocean. However, a comparison based on the NASA GEOS-1 atmospheric model shows that the wind frictional torque obtained is insufficient in power in the x and y components. Finally, in the application to the Earth rotation excitation it is found that both the ellipticity and topographic torques have comparable contributions to polar motion excitation for both seasonal and nonseasonal variations. For the length of day variation, our torque approach is presently inadequate because the oceanic contribution is relatively too small to be estimated in the presence of data and model errors.