Atmospheric records of sulfate and MSA in aerosols were obtained by year-round samplings in the boundary layer at two coastal Antarctic sites, Neumayer (1983--1995) and Dumont D'Urville (1991--1997). At Dumont D'Urville and Neumayer the non-sea-salt sulfate (nssSO42-) level observed from October 1991 to October 1993 exceeded by 44 and 48 ng m-3, respectively, the mean level (~150 ng m-3) observed during nonvolcanic years. The Neumayer and Dumont d'Urville records first exhibit spring (October to December) 1991 nssSO42- concentrations exceeding mean spring concentrations by some 100 to 150 ng m-3 caused by a significant poleward transport of the Cerro Hudson volcanic aerosol in the lower stratosphere beneath the polar vortex and possibly in the middle/upper troposphere. Limited to 10 ng m-3 during winter 1992, the nssSO42- perturbation reached 78 ng m-3 in fall 1993 at Neumayer. The latter enhancement corresponds to the input of the Pinatubo volcanic aerosol advected through stratosphere/troposphere air mass exchanges. An enhancement of nssSO42- concentrations is also recorded in snow layers deposited between 1991 and 1993 at various sites located over the high Antarctic plateau. On the basis of these firn data the relative impact of the three latest volcanic eruptions of global concern, namely, Mount Agung (1963), El Chichon (1982), and Pinatubo (together with Cerro Hudson) (1991), was evaluated in terms of sulfate input at high southern latitudes. The fallout of sulfate in central Antarctic snow following the Cerro Hudson/Pinatubo eruptions was slightly lower than fallout after the Mount Agung eruption. Sulfate fallout after El Chichon was 2 to 3 times lower than that from either Mount Agung or Pinatubo. The enhancement of sulfate levels detected in central Antarctic snow deposits during the 1991--1993 time period allows us to estimate a mean atmospheric sulfate perturbation of around 60 ng m-3 in the boundary layer of these regions (i.e., at least 35% higher than that seen at coastal sites). This observation suggests that coastal Antarctic regions are less sensitive than the high plateau in detecting volcanic sulfate fallouts following eruptions of global concern. Atmospheric near-surface, year-round sampling achieved at coastal Antarctic sites' reveals no significant enhancement of the nitrate level over 1992 and 1993, suggesting that volcanic aerosols have had a limited effect on the transfer of nitric acid from the lower Antarctic stratosphere to the boundary layer in these regions. ¿ 1999 American Geophysical Union