We determine subsidence in the Arctic stratospheric vortex from measurements of the column density of hydrogen fluoride, using a rotational line thermal emission. The data were obtained with the far-infrared spectrometer on board the NASA DC-8 aircraft during the Arctic Airborne Stratospheric Expedition, from January through March 1992. A wide range of meteorological conditions was sampled, both inside and outside the polar vortex region. The measured values of subsidence are analyzed as a function of time of year and potential velocity. From this analysis we draw the following conclusions: (1) Subsidence is strongly correlated with potential vorticity. However, there appear to be other factors, including meteorological conditions, which play a lesser role. (2) The gradient of subsidence with respect to potential vorticity is large and approximately constant across the vortex wall and is small elsewhere. The gradient of subsidence with respect to horizontal distance is dramatically large across the vortex wall. (3) There appears to be substantial descent in the early winter vortex, from the evidence of the observed subsidence in January and the assumption of zero subsidence at the fall equinox. (4) Descent continues to occur in late winter, with measured vertical velocities similar to that inferred for early winter. The averaged early and late winter vertical velocity at a representative altitude of 18 km is 0.052(¿0.013) cm s-1. (5) At the measured rate of subsidence the time required to flush the stratosphere by one scale height is about 6 months. ¿ American Geophysical Union 1995