Sulfur stable isotopes from Antarctic snow samples have been used to assess sources of sulfate. The novel experimental procedure presented here is suitable for the determination of sulfur isotopic composition at the micromolar level and has been adapted to polar ice samples. Measurements were carried out on three contiguous firn cores (PS6, PS7, and PS8) collected near Amundsen-Scott Station (South Pole), covering the record of the Agung eruption (March 1963). Taking into account the minimum amount of sulfate required for the isotope analysis, it has been possible to delineate three time periods along the cores: pre-1964 years (background sulfate level), 1964--1965 (volcanic deposition peak), and 1966--1968 (volcanic peak tail). A deeper part of another core (PS12) has been used to extend the background picture. Assuming the conservation of isotopic signatures during long-range transport and deposition processes, results demonstrate the significant volcanic contribution to sulfate deposition on the central Antarctic ice cap a few months after a major low-latitude eruption. They also confirm the marine biogenic origin of present background sulfate. Isotopic signatures (Δ34S) of marine biogenic sulfate and volcanic sulfate from Mt. Agung have been found to be +18.6¿0.9? and +2.7¿1.1?, respectively. ¿ 2000 American Geophysical Union