Nevado del Ruiz volcano has been releasing extraordinarily large volumes of sulfur dioxide gas from its Arenas crater since September, 1985 and probably since the renewal of volcanic activity began in late 1984. The combined release from eruptive and passive flux has been approximately (3.4 ± 1.0) × 10 6 metric tons in the four years after November, 1984. This value combined with data of others on pre-eruption sulfur content of the magma, allows the calculation of a minimum required volume of magma of 0.92 km 3. Consideration of the continued high flux and typical ratios of erupted to degassed magma indicate that the true minimum volume of magma involved is probably at least 4.6–9.2 km 3. A systematic study of the various species of sulfur in the volcanic plume and the hydrothermal system has provided a characterization of the geochemistry and sulfur-isotopic variation. Consideration has been given to five potential sources of the sulfur dioxide: dissolution of subvolcanic evaporites; remobilization and oxidation of native sulfur within summit crater fumaroles; large-scale release from the hydrothermal system by reduction of sulfate or water-rock interaction; assimilation and oxidation of a pre-existing sulfide deposit; and magmatic volatiles. The data are most realistically explained by the passage of magmatic gas through the hydrothermal system, with some disproportionation of sulfur dioxide to produce the high sulfate content and low pH of the waters. This model is also most consistent with the distribution of acid-sulfate-chloride hot springs, high chloride and fluoride content of the hydrothermal waters, sulfide deposition within the volcanic pile, and the transient increase in sulfate content of the acid-sulfate-chloride hot springs that reached a maximum approximately one year after the November, 1985 eruption. The magmatic model is in apparent conflict with the absence of general deformation from the time that data began to be collected, about one week before the eruption, until the time this manuscript was finalized (September, 1989). Although this conflict cannot be resolved with certainty, it is noted that the absence of deformation is only an indication of the lack of magma movement, not of the absence of magma. An alternative magmatic model considers the available geological data and proposes that the system is related to a large-scale caldera. In this model, the degassing is related to a long-term process of exsolution of magmatic volatiles from a large volume of crystallizing magma left from the construction of modern Ruiz. This model can better explain the combined geological and geophysical data, as well as the geochemical and isotopic data. The nature of the degassing data at Ruiz is interesting because it is an example of an explosive eruption cycle in which the gas flux is actually dominated by passive release. The nature of the degassing allows us to hypothesize that some major change in the system occurred in April–May, 1986, and to predict that the present high flux can be expected to last for another decade.