Projections of Antarctica’s contribution to sea-level rise are highly uncertain due to our limited understanding of the processes driving ice-mass changes in a warming climate. While the ice-dynamically induced hotspots of mass loss are known, a long-term increase in surface mass balance (SMB) is poorly constrained. This is because long-term climate trends in SMB are subtle and masked by the large natural SMB variability. In this study, we explore advanced methods to resolve long-term trends in SMB. We do this by jointly analysing elevation changes from satellite altimetry and from SMB and firn modelling in space and time back to 1992. By synergistically combining both data sets, we quantify the large interannual variations related to SMB and firn processes, allowing a robust estimation of the underlying long-term trends. In order to assess a realistic trend uncertainty and trend significance, our approach accounts for temporal correlations in the altimetry time series due to both SMB-related time-variable signals and altimetry errors such as radar penetration. The analyses are carried out at the level of monthly grids with a spacing of a few kilometres. The resulting high-resolution spatial trend patterns enable interpretation and attribution to SMB-related or ice-flow related processes.