A dynamic-thermodynamic sea ice model is used and coupled with an atmospheric general circulation model to simulate the seasonal cycle of the global sea ice distribution. We have run the coupled system and obtain a creditable seasonal simulation of the Antarctic sea ice. To understand the role of ice advection on the seasonal cycle of Antarctic sea ice in the coupled system, results from the thermodynamiconly (T) sea ice model have been compared with those from the dynamic thermodynamic (DT) sea ice model. The seasonal cycle of sea ice differs between the two models. When ice motion is eliminated sea ice becomes more compact and thinner, and sea ice is more extensive in summer. A number of previous studies have examined the effect of ice dynamics on sea ice simulations with prescribed atmospheric conditions. Here experiments have been performed with a fully coupled atmosphere sea ice system and also using prescribed daily atmospheric forcing and monthly mean atmospheric forcing, to examine the differences of the sensitivity of the ice advection between the coupled and forcing models. Similar differences have been observed between DT and T in the forcing models but the magnitude is smaller than in the fully coupled model, and with monthly mean atmospheric forcing the difference is least. These differences highlight the importance of the inclusion of ice advection when undertaking studies using a fully interactive atmosphere sea ice model, or using prescribed daily/monthly atmospheric conditions to force a sea ice model for the Antarctic.