In this paper, an algorithm is developed based on the split-window technique, to estimate the SO2 abundance in the plume of Mt. Etna volcano using the multispectral infrared and visible imaging spectrometer (MIVIS). The MIVIS data were remotely sensed in the thermal infrared (TIR) during the Sicily-1997 Campaign. In this study, the MODTRAN 3.5 code has been used to simulate the radiance at the sensor; the radiative transfer model was input along with the data of radio-sounding performed simultaneously with the MIVIS flight using a mobile radio-theodolite. From the SO2 map, derived from the MIVIS image, the SO2 flux along the axis of the plume was computed knowing the wind speed at the plume altitude. The SO2 flux is variable along the plume axis. The average SO2 flux (about 45 kg s–1 on 12 June and about 30 kg s–1 on 16 June) emitted from the vents is compared with the correlation spectrometer (COSPEC) measurements carried out by other teams (from the ground and from a light aircraft flying under the plume) during the MIVIS flight. Finally, by means of this algorithm it should be easier, with respect to the previously described procedure to monitor the SO2 flux of a specific volcano such as Mt. Etna.