We studied how gravity data can be used to estimate the shape of the seafloor beneath Antarctica’s floating ice shelves, where direct measurements are difficult. Using computer models based on real data, we tested when this method works well and where it has limits. We found that it could greatly improve seafloor maps for most ice shelves with high-quality gravity data. Better models of the seafloor will help us understand how ocean water melts ice from below, affecting future sea level rise.