Abstract: We consider trends of sea level steric oscillations in the North Atlantic in 2003–2015 using two independent approaches. The first method is based on the multidisciplinary application of altimetry (AVISO) and gravimetry (GRACE mission) data. The second method is based on the integral assessment of steric oscillations using variations in the fluid volume caused by variations in water density: the calculations are carried out using the SODA, EN4, and ARMOR reanalysis data. It is shown that the application of the combined altimetry and GRACE data result in overestimated values of steric oscillations and their trends. This is related to the fact that the GRACE observations show the variations in the ocean mass; hence the sea level variations are presented in a relative coordinate system which is not the geocentric system. This system does not take into account the effects of the elastic deformation of the ocean bottom and the corresponding redistribution of the water volumes. It is shown that the maximum bias of these estimates and the errors in determining the steric oscillations and their trends based on the first method is characteristic of regions located near Greenland. This is caused by the contribution of the negative trend component to the GRACE data. If the estimates are made in regions remote from Greenland coasts, the trend component in the GRACE measurements is insignificantly manifested and the trends in the steric oscillations calculated using the method that jointly uses the AVISO and GRACE data are similar to the trends in the sea level variations based on the altimetry data. The trends of the steric oscillations of the sea level calculated from the reanalysis data are similar in the spatial distribution between both calculations and also similar to the trends in the sea level variations based on the altimetry data.