Annual and interannual variability of 21 years of passive microwave brightness temperatures on the Antarctic ice sheet is documented through principal component analysis. The leading modes show that brightness temperatures are dominantly forced by the annual temperature cycle, with surface melting signals explaining about 5% of the total variance of the data. Analysis of the data with the annual cycle and melting signals removed results in two significant interannual modes. The first is consistent with the Antarctic Oscillation and its influence on Antarctic surface temperatures, while the second reflects regional scale variability that appears to be linked to sea-ice anomalies and the El Ni¿o-Southern Oscillation. The expression of these characteristic patterns of Southern Hemisphere climate variability on the Antarctic continent provides a framework for interpreting modern climate records, as well as a basis for reconstructing past climate variability with ice cores.