Combining SO2 plume emissions from Mt. Etna in 1975-1995 with the magma S content suggests that between 3.5 and 5.9 km(3) of basalt were degassed for sulfur over two decades, only 10-20% of which actually extruded. Except during extensive lava outbreak (1992), the SO2 flux was mostly supplied by endogenous degassing of non-erupted basalt, replaced at a time-averaged rate of 4.5-8 m(3)/s. The unerupted degassed magma cannot be accomodated by the upper plumbing system (whose maximum capacity is estimated at 0.6 km(3)), nor by simple intrusive growth of the volcanic pile, Being denser than undegassed melt, most of it was probably removed by gravitational convection in subvolcanic feeders, Ultimately, part of this degassed magma may solidify within the crust, contributing to the accretion of a wide 'plutonic' complex, greater than or equal to 3 times larger than the volcano itself, that has grown within its sedimentary basement.