This modeling study was motivated by the recent publication of year-round records of dimethylsulfide (DMS) and dimethylsulfoxide (DMSO) in Antarctica, completing the available series of sulfate and methanesulfonic acid (MSA). Sulfur chemistry has been incorporated in the Laboratoire de M¿t¿orologie Dynamique-Zoom Tracers (LMD-ZT) Atmospheric General Circulation Model (AGCM), with high-resolution and improved physics at high-southern latitudes. The model predicts the concentration of six major sulfur species through emissions, transport, wet and dry deposition, and chemistry in both gas and aqueous phases. Model results are broadly realistic when compared with measurements in air and snow or ice, as well as to results of other modeling studies, at high- and middle-southern latitudes. Atmospheric MSA concentrations are underestimated and DMSO concentrations are overestimated in summer, reflecting the lack of a DMSO heterogeneous sink leading to MSA. Experiments with various recently published estimates of the rate of this sink are reported. Although not corrected in this work, other defects are identified and discussed: DMS concentrations are underestimated in winter, MSA and non-sea-salt (nss) sulfate concentrations may be underestimated at the South Pole, the deposition scheme used in the model may not be adapted to polar regions, and the model does not adequately reproduces interannual variability. Oceanic DMS sources have a major contribution to the variability of sulfur in these regions. The model results suggest that in a large part of central Antarctica ground-level atmospheric DMS concentrations are larger in winter than in summer. At high-southern latitudes, high loads of DMS and DMSO are found and the main chemical sink of sulfur dioxide (SO2) is aqueous oxidation by ozone (O3), whereas oxidation by hydrogen peroxide (H2O2) dominates at the global scale. A comprehensive modeled sulfur budget of Antarctica is provided.