The possible impact of anthropogenic climate change on the Asian summer monsoon is investigated in several time-slice experiments using prescribed sea-surface temperature (SST) and sea-ice anomalies. The study is carried out with four different atmospheric general circulation models (GCMs), each being involved in two pairs of experiments differing only by the treatment of the land surface hydrology. The objective is to assess the robustness of the simulated climate change, and its possible sensitivity to the land surface scheme. Despite the use of identical SST anomalies, the four GCMs do not predict similar monsoon responses on the regional scale. All models produce a stronger warming over the Asian continent than over the Indian Ocean, but this warming is not a good predictor of the monsoon response to increased CO₂ level. There is a significant spread in the summer precipitation anomalies despite a general weakening of the monsoon circulation, showing that the response of the monsoon rainfall is not solely related to the changes in the large-scale dynamics. In a warmer climate, the monsoon precipitation can increase despite a weakening of the monsoon flow, due to an increase in the atmospheric water content. For decades to come, the increase in the atmospheric water content could be more important than the increase in the land-sea thermal gradient for understanding the evolution of the monsoon precipitation. Though it does not represent a major source of uncertainty, the treatment of the surface hydrology is liable to affect significantly the regional response of the monsoon to CO₂ doubling. A slight change in evapotranspiration is enough to induce a significant change in precipitation. A simple analysis of the regional water budget indicates that this sensitivity is not only related to changes in the horizontal transport of water vapor, but also to changes in the precipitation efficiency, which depends on the treatement of the land surface hydrology.