In unpolluted areas sulfur occurs primarily in three compounds, SO4= in aerosols and SO2 and H2S as gas. Sources of SO4= are the ocean, the soil, and the oxidation of the gaseous compounds. A considerable fraction of the SO2 is of anthropogenic origin. As measurements in polluted and unpolluted areas indicate, most of the H2S seems to be natural. In unpolluted air in the northeastern part of the United States, both gases have approximately equal concentrations. The chemistry of the conversion of these gases to SO4= is discussed. Estimates of the sulfur budget of the global atmosphere, excluding the SO4= in sea salt particles, indicate that 30 per cent is of anthropogenic origin. Measurements of the chemical composition of Greenland ice back to the year 1915 are presented. The measurements show that the concentration of SO4= is much higher than that of other soluble constituents, but it has not increased since 1915 as much as one would expect if it were of industrial origin. It is concluded that industrial sulfur is removed from the atmosphere before it spreads to polar regions. This is consistent with estimates of the residence time of about 5 days for SO2 and about 40 days for total sulfur. Convay showed that more SO2 enters the oceans with river water than can be accounted for by the weathering of rocks and sediments on land and that the excess must come from the atmosphere. He assumes a closed sulfur cycle and suggests that this sulfur must return to the atmosphere as H2S, produced in the shelf areas of the oceans. In view of the considerable amounts of industrial sulfur, it is possible that the global sulfur cycle is not closed and that most of the excess sulfur in river waters is of anthropogenic origin and will stay in the ocean.