Sea ice freeboard heights in the Weddell Sea of Antarctica are derived from the Ice, Cloud, and Land Elevation Satellite (ICESat) laser altimeter measurements, which have a unique range precision to flat surfaces of 2 cm within 70 m footprints spaced at 172 m along track. Although elevations of flat surfaces can be obtained to an accuracy of ∼10 cm (1σ) per footprint, direct determination of freeboard heights is precluded by errors in knowledge of the geoid and temporal variability of the ocean surface. Therefore freeboards are determined relative to an ocean reference level detected over areas of open water and very thin ice within the sea ice pack using an along‐track filtering method. The open water/thin ice selections show good agreement in the combined analysis of ICESat segments and Envisat Synthetic Aperture Radar (SAR) imagery. The average residual between the ICESat‐measured ocean level and the EGM96 geoid is 1.4 m. Estimates of snow depth on the sea ice from AMSR‐E passive microwave data along with nominal densities of snow, water, and sea ice are used to estimate sea ice thickness. Four periods of ICESat data in May–June (fall) and October–November (late winter) of 2004 and 2005 between longitudes 298°E and 360°E are analyzed. In the fall the mean freeboards are 0.28 m in 2004 and 0.29 m in 2005, and the mean thicknesses are 1.33 m in 2004 and 1.52 m in 2005. In late winter the freeboards grew to 0.37 m in 2004 and 0.35 in 2005, and the thicknesses grew to 2.23 m in 2004 and 2.31 m in 2005. The interannual differences in freeboard are small, and the larger interannual change in estimated thickness mainly represents differences in the snow depth estimates. Seasonal changes in the spatial patterns of freeboard and thickness over the 4 months correlate with the expected circulation of sea ice in the Weddell Sea, as indicated by sea ice velocity fields.