It has been recognized since the commencement of Antarctic ozone measurements during the IGY that spring southern polar total ozone amount is less than spring northern polar total ozone amount. More importantly, since 1980 there has been a decline in the minimum spring total ozone value, from 250 DU in 1980 to 125 DU in 1987 and below 120 in 1991. This decline occurs within the winter polar vortex, which acts as a containment vessel preventing polar ozone from escaping to lower latitudes and excluding ozone-rich air from the polar region. Ozone decrease can be explained in terms of heterogeneous reactions of chlorine and nitrogen reservoir molecules on polar stratospheric clouds. These clouds form in the lower polar stratosphere during winter when temperatures in the Antarctic are sufficiently low to create water ice clouds. Clouds involving nitric acid form at higher temperatures. Chlorine reservoirs such as HCl are converted to Cl₂, which is photodissociated in the presence of sunlight. The resulting Cl reacts with O₃ to form ClO. Measurements of ClO and other species give agreement of theory and experiment within the uncertainties of the measurement. Heterogeneous chemistry accounts for most of the ozone hole. A small amount of ozone loss is also observed above the polar stratospheric cloud level, implying another mechanism, either chemical or dynamical. Above 25 km, formation of ozone-destroying odd nitrogen in the upper stratosphere by energetic electrons and the existence of any trend is still an open question. There is much less ozone depletion in the Arctic. This is the result of a less stable polar vortex and warmer temperatures, which reduce polar stratospheric cloud formation. There is strong evidence that tropospheric forcing within or just outside the vortex leads to adiabatic cooling with resulting cloud formation. During such events ozone-poor tropospheric air is transported into the stratosphere. In the Arctic this can result in the transport of long-lived hydrocarbons. Subsequent reactions lead to the formation of HCl, reducing the effect of Cl. There is also production of HO₂, which accelerates ozone loss due to chlorine. There are also small areas of large and rapid ozone depletion termed miniholes. Ozone-poor air from these regions can propagate to lower latitudes, as can the air from within the vortex, when it disintegrates in late spring. Data from the BUV ozone-measuring instrument on the Nimbus 4 satellite indicate the existence of October 1970 Antarctic ozone of only 250 DU. This is evidence of the existence of ozone loss with only CH₃Cl and low concentrations of CFCs as chlorine sources.