Possible future changes in land ice volume are mentioned frequently as an important aspect of the greenhouse problem. This paper deals with the response ofthe Antarctic ice sheet and presents a tentative projection of changes in global sea level for the next few hundred years, due to changes in its surface massbalance. We imposed a temperature scenario, in which surface air temperature rises to 4.2¡C in the year 2100 AD and is kept constant afterwards. As GCMstudies seem to indicate a higher temperature increase in polar latitudes, the response to a more extreme scenario (warming doubled) has also beeninvestigated. The mass balance model, driven by these temperature perturbations, consists of two parts: the accumulation rate is derived from presentobserved values and is consequently perturbed in proportion to the saturated vapour pressure at the temperature above the inversion layer. The ablation modelis based on the degree-day method. It accounts for the daily temperature cycle, uses a different degree-day factor for snow and ice melting and treatsrefreezing of melt water in a simple way.According to this mass balance model, the amount of accumulation over the entire ice sheet is presently 24.06x1011 m3 of ice, and no runoff takes place. A1¡C uniform warming is then calculated to increase the overall mass balance by an amount of 1.43x1011 m3 of ice, corresponding to a lowering of global sealevel with 0.36 mm/yr. A temperature increase by 5.3¡C is needed for the increase in ablation to become more important than the increase in accumulation andthe temperature would have to rise by as much as 11.4 ¡C to produce a zero surface mass balance. Imposing the Bellagio-scenario and accumulating changesin mass balance forward in time (static response) would then lower global sea level with 9 cm by 2100 AD. In a subsequent run with a high-resolution 3-Dthermomechanic model of the ice sheet, it turns out that the dynamic response of the ice sheet (as compared to the direct effect of the changes in surface massbalance) becomes significant after 100 years or so. Ice-discharge across the grounding-line increases, and eventually leads to grounding-line retreat. This isparticularly evident in the high scenario and is important along the Antarctic Peninsula and the overdeepened outlet glaciers along the East-Antarctic coast.Grounding-line retreat in the Ross and Ronne-Filchner ice shelves, on the other hand, is small or absent.