The tropospheric sulfur cycle has been closely studied from the standpoint of anthropogenic emissions but less so with respect to volcanic sources. We document here evidence for wide variation in lifetimes of volcanic SO2 in the troposphere. At one extreme, our observations of the plume associated with the lavadome eruption of Soufriere Hills volcano (Montserrat) suggest loss of gaseous SO2 at rates exceeding 10-3 s-1. While this efficient SO2 depletion reflects unusual environmental factors, published measurements for Mount Etna's plume approach these rapid rates. In such cases, source strengths of SO2 must significantly exceed apparent fluxes measured several kilometres downwind. This implies that meteorological and geographic factors, as well as volcanic degassing rates, should be considered in the interpretation of fluxes of SO2 measured beyond source vents, with ramifications for volcano monitoring and eruption prediction. Furthermore, since most calculations of global volcanic fluxes of sulfur (and other species) to the troposphere are based on extrapolation of SO2 flux data they may underestimate true source strengths. Finally, we consider that fast sulfur chemistry may also prevail in convecting eruption columns, resulting in partial tropospheric scrubbing of sulfur from stratosphere-bound plumes. ¿ 1998 American Geophysical Union