The depth to the upper surface of an 'ideal' South Pacific oceanic lithosphere is calculated by applying a well-known thermal subsidence equation to a grid of ocean floor ages that is generated from published isochrons interpreted from magnetic and gravity anomaly data. When these depths are removed from observed bathymetry, a grid of residual bathymetry results, which at long wavelengths can be related to the presence of seafloor sediments. By employing the concept of an isostatic correction for sedimentary loads, which describes the depth difference between observed bathymetry of a sedimented ocean and the thermal subsidence surface, it is possible to model sediment thickness using the residual bathymetric anomaly. Long wavelength uncertainty in this thickness may be related to dynamic topography and/or large scale crustal thickness variations, and becomes larger with increasing residual topography; with data uncertainties included the maximum possible errors in sediment thickness estimates are in the range 3.8-6.9 km. Local flexural effects, abyssal hill topography, and crustal thickness variations due to seamount volcanism are likely to give rise to further uncertainty at short and medium wavelengths. At the West Antarctic margin, sediment volumes estimated in this way are larger than those based on isopachs constructed from seismic data. This may be related to inadequacy in seismic data on which the isopachs are based, and/or to dynamic topography and crustal thickness variations.